The hand crane by the same patentee had a lifting power of 3 tons. The steam winch (Mégy's patent) was the most efficient and economical of all the hoisting appliances. Full details of this and all the other machines employed are given by the Author in the plates which accompany the Paper.

The statical computations were made for two continuous braced girders with parallel straight booms, compression and tension bars, and two footways each 61 feet wide. The clear width of each opening is 308 3 feet, the effective width 321 4 feet, and the depth of the girder one-tenth of this latter dimension. The computations are made under the assumption that the loads on the bridge per lineal foot was 1.38 ton (4.56 ton per lineal mètre) for fixed loads, and 1·58 ton (5.21 tons per lineal mètre) for moving loads, and that the moving load may be considered as uniformly distributed; that the cross section of the girder is constant; that the limiting stress may be considered constant, and without regard to the differences of the strains, and to the effective width; and that the fourfold system of the bars could be split up into four single systems. The Author shows at great length the admissibility of these assumptions.

The booms of the main girder were box-shaped, the horizontal plates of which were 4 feet wide, and attached to bars from 1 foot to 2 feet wide. These plates were made to abut as nearly as possible at the joints, and the stays and angle irons were butted in the middle of each bay. At the joints there were special plates for connecting the tension bars, of which there were four kinds. The compression bars were attached outside the stays, and every pair was connected with some part of the structure by a flat-bar lattice bracing.

The vertical ribs were also connected by a similar lattice bracing, and they served to secure the upper cross stays and cross girders, and to support the footway cantilevers. The cross and longitudinal girders and the wind stays or braces were of the usual construction. The saddles of the free movable supports were of cast iron, and the rockers of cast steel, lead sheeting being laid between the boom and the upper saddle plate, and also between the lower plate and the top of the pier below. All the ironwork was made by Cail and Co., of Paris, the different parts being there riveted together in convenient lengths, and simply fitted at the site of the bridge. Detailed drawings are given of the machines used to test the strength of the materials employed, the results of which are recorded in tables.

The setting-up of the girders was done on a fixed staging, and at first was carried out in short lengths of two or three bays on each side of a pier; but as it was thus difficult to insure the exact connection and position of the lower boom, a modification was made, and the lower booms of No. 3 pier were set out at their full length on the staging; the cross and longitudinal girders fixed between them, and rails laid on the latter, on which travelled

cranes, winches, and wagons, for the lifting up and conveyance of the material; the verticals and struts were then set up in position, and the upper boom attached.

In July 1875 a violent hurricane occurred, which moved the whole girder of the second span (just then completed) more than 3 inches out of position, and the girders of the third span were similarly displaced by a gale in October of the same year. Their restoration to the original positions, work of difficulty and anxiety, was effected by American screw-jacks and winches. The cost

of sinking and laying the foundations was about 28. 2d. per cubic foot for the shore piers, and from 18. 10d. to 2s. 1d. for the river piers.

The average rates for masonry were as follows :

:

From the drawings it appears that the total length of the girders was 1,290 feet, but the leading dimensions have not been given in the paper. The weight of metal in the bridge is over 3,000 tons. W. H. E.

The St. Louis Bridge over the Mississippi. By M. LAVOINNE.

(Annales des Ponts et Chaussées, 5th series, vol. xiv., pp. 5–71, 2 pl.)

The convergence of several railways at St. Louis, and on the opposite bank of the Mississippi, rendered it necessary to erect a bridge over the river to connect them. A company was formed for the purpose, and incorporated by Act of Congress in 1868. Peculiar difficulties had to be encountered in laying the foundations of the bridge, for the sand forming the bed of the river, which is largely deposited in the summer, is liable to be scoured away to a great depth when the river is coated with ice, owing to the contracted undercurrent acquiring a great velocity. The dip of the underlying rock is also very great, being only 13 feet below lowwater level on the St. Louis side, whilst on the opposite side it is 108 feet. Under these circumstances it was absolutely necessary to carry the foundations down to the rock, and as it was important to reduce the number of the piers as much as possible, both on account of the costly nature of the foundations and of the shortness of the suitable season for erecting them, being only from the middle of August to the middle of December. The site chosen was a narrow part of the river opposite the centre of the town; and the material selected for the superstructure was cast steel, so that the

spans might be made very large. The bridge was designed to have three spans, and, as it had been found by experiment that cast steel could bear a considerably greater strain in compression than tension, girders of two arched steel tubes, placed one above the other and connected by bracing, were adopted.

The western abutment was commenced in 1868, and the foundations, 13 feet below low water, were laid by means of an ordinary cofferdam. The foundations of the two piers and of the eastern abutment were carried down to the rock by means of caissons sunk by the aid of compressed air, and a hydraulic sand pump was used for removing the sand from the bottom of the caissons. The caissons for the piers were made of plate iron, of a hexagonal form, 82 feet long and 60.7 feet wide, divided into three compartments, each 19.7 feet wide. The working chamber, 9 feet high, was connected with the upper part of the caisson by seven shafts, the air locks having been placed at the lower end of the shafts to avoid removing them each time that a length was added to the caisson, and rendering the working chamber the only part needing to be filled with compressed air. The sinking of the caisson of the east pier to a depth of 102 feet below low water occupied one hundred and thirty-three days, and filling the working chamber with concrete fifty-three days. For the west pier these operations occupied seventy-seven and fifty-three days respectively, the depth being 76 feet. The bottom and sides of the caisson of the eastern abutment were made chiefly of wood instead of iron, with layers of oak beams above the working chamber to a thickness of 5 feet. The caisson was made watertight by a casing at the sides, of plate iron 3 inch thick, and over the top of the working chamber, and projecting 10 inches below the bottom to assist the sinking. There were three shafts, one connected with a double air lock, and the two others with single ones placed at the level of the working chamber. The caisson was sunk to a depth of 108 feet, and only the outer envelope of the working chamber was filled with concrete, the interior being filled with sand. The spans of the arches, measured at the level of the springing of the upper tube, are 520 feet for the centre opening, and 502 feet for the two side openings, with a rise of 59 feet 8 inches and 56 feet 5 inches respectively. The lower roadway of the bridge for carrying the railway is placed at the level of the top of the underside of the arches, and the carriage roadway is 261 feet above the railway at the centre of the bridge. The width between the parapets is 54 feet. The tubes are 12 feet apart centre to centre. The distances between the four girders are 12 feet for the centre ones and 16 feet for the side ones. The piers at the springing of the arches are 32 feet wide. The steel tubes, forming the top and bottom flanges of the segmental trellis girders, have a mean length of 11 feet 10 inches, and an external diameter of 1 foot 6 inches, and are connected by being grooved into collars joined by bolts. The tubes consist of six segments of laminated steel, varying in thick

ness from 14 inch to 2 inches from the centre to the springing, by variations of the internal diameter, and fastened together by a steel casing inch thick, which butts at its ends against the connecting collar, and is pressed against the segments by bolts. The collars are pierced horizontally by stout steel pins with cylindrical ends projecting beyond the collars, on which wroughtiron bars fit, bracing together the two tubes. These bars vary in width between 1 foot 1 inch and 1 foot 8 inches, and in thickness between 1 inch and 2 inch from the centre to the springing; and, as they are liable to be subjected to both tension and compression, each couple is braced together. The lower roadway, carrying two lines of rails, each line passing between one central and one side girder, is supported by cross girders fastened to bars hanging down from the upper tube where they are below it, and to the upright bars supporting the upper roadway for the rest of the span. The upper roadway is borne on cross girders, which rest on the upright bars fastened to the upper tube. The space between the cross girders is covered with iron plates, inch thick, which serve to protect the roadway from sparks from the engines below, and to stiffen the bridge against gales of wind. Wind ties are placed in the portions of the arches below the railway connecting the bottom tube of one girder to the top tube of the adjoining one, and also above the railway between the two central girders. The limit of the safe strain was estimated at 12.7 tons per square inch of section in compression and 8.9 tons in tension.

The erection of scaffolding for putting the girders in place would have been both difficult and costly, on account of the shifting nature of the river bed, the depth of the solid rock, and the amount of protection needed against damage in floods or in the breaking up of the ice. The navigation also would have been impeded; and the progress would have been slow, as only the scaffolding for one span could have been erected at a time. The girders were accordingly erected from the piers and abutments by joining together a quarter of the span length by length, which, fastened to the piers at the springing, could bear its own weight like a bracket. Then by supporting this portion from above by a system of iron bars, carried over a timber tower erected on the pier, and fastened to a similar length on the opposite side, or at the abutments, anchored to the bank, some more lengths of tubes and bracing could be added, and the end of this additional portion supported by bars, passing over a mast raised up on the girder and joined at the opposite end to the springing plate, which enabled the central portion to be added. When the two central girders were in place, the outer ones were put together from scaffolding put out from the central girders. The weight of a girder over the centre opening is 217 tons, of which about two-thirds consists of steel. If iron had been used instead of steel the weight would have been two and a half times greater, and the increase in expense £26,000.

The total cost of the bridge, and of the works for connecting it with the railways on each shore, is as follows::

The total length of the roadway of the bridge proper is 1,300 feet, and that of the approaches 534 feet; together 1,834 feet. The bridge was opened for traffic in July 1874.

L. V. H.

Cable Making for Suspension Bridges, as exemplified at the East River Bridge, New York.

By W. HILDENBRAND.

(Van Nostrand's Engineering Magazine, vol. xvii., pp. 171–184, 193–201, 289–300.)

The method adopted by the Chief Engineer, Col. Roebling, for the cables of the East River Bridge is the same as that used in the case of the Niagara and other suspension bridges, and consists in taking each wire of a cable over the opening by itself, then adjusting its length, and finally combining them all by a wrapping into a cable. This was preferred to that of making a cable of wire ropes, mainly because the strength of a straight wire is 10 per cent. greater than that of a twisted one: hence a wire-rope cable would have been heavier, more costly, and more bulky, exposing a greater surface to wind and atmospheric influences. It was also preferred to the third possible method of using straight wires, but first making up the cable on shore, and then hauling it into position, because this requires a space equal in length to the cable to be available on one or the other side of the bridge (since straight wires cannot be coiled up for transportation), and this, in such a crowded district, was out of the question. In addition, where a cable is made up first and suspended afterwards, the upper strands are strained more than the lower, and thus the distribution of the load cannot be accurately known. The disadvantage of the method adopted is