The last specimen broke under a strain of 84,032 lbs. per square inch, and showed a beautiful fracture entirely fibrous.

The cast-iron used in the details was a gray iron formed of a mixture of pig, generally adopted by the Keystone Bridge Company.

Specimens were tested by suspending a weight upon a bar two inches by one, and placed upon support four feet apart. The specifications required that this breaking weight should not be less than 2,100 pounds, and in all of the tests it was found to be much in excess of this amount.

The shore span is a riveted trellis girder of wrought-iron 71 feet long and 8 feet deep. The chords are of T section, composed of two vertical plates, one horizontal plate and two angle pieces; in the bottom chord the horizontal plate does not reach to the ends of the span, and the other parts are continuous for the whole length. The braces are each formed of two pieces of Ţ iron placed back to back, and enclosing the ties, which are single bars of flat iron; both ties and braces are riveted between the vertical chord plates. The laterals are of wrought-iron, and the trusses are stiffened by short braces of iron connecting the floor with the web. The end posts are enclosed in light ornamental castings. The floor beams are of pine, six inches by eighteen, without trussing, placed two feet between centres; on this is laid a floor similar to that on the other fixed spans. The amount of material in this span, exclusive of pavement and hand rail, is as follows: Lumber, 7,684 feet B. M.; wrought-iron, 32,165 pounds; cast-iron, 4,328 pounds.

The draw measures 361 feet and 3 inches over all; it is a Pratt truss of similar plan to the large draws erected by the Keystone Bridge Company at Cleveland, Dubuque, and other points. The skew is taken out of the truss by making the end panels of unequal lengths, the difference being 5 feet 6 inches. The upper and lower chords are of like pattern, formed of two I beams and two channel bars eight inches deep, placed side by side and united by a plate riveted to their upper flanges. The posts are of wrought-iron, of the Linville pattern. The ties are round, with both ends upset for screws; the main ties are in pairs, and the counters single, passing through the posts. The washer plates upon which bear the nuts of the ties are of cast-iron, except the top centre, which is forged. The floor beams are ten-inch rolled I beams, and rest on

the top of the lower chord. The floor is of two-inch oak plank laid on the oak track stringers, and pine floor joists. There is no separate footway on the draw. The turn-table is formed of an external drum thirty feet in diameter, and a central shell of cast-iron, hung by ten bolts on one of Sellers' patent pivots; the drum and shell are connected by a pair of plate girders under the centre posts, and a set of radial rods. The bolts are adjusted so as to throw almost the entire weight on the centre, the drum serving only as a guide and balancer. The draw is easily opened by four men, with leve:s attached to two pinions on the drum, in two minutes, but as a precaution against wind and other dangers, it is to be fitted with a steam-engine. The latch is worked from the centre by a hand-lever; a bearing is secured by wedges which are driven under the four end-posts, the four being worked by a single central lever. The amount of material in the draw, including both trusses and turn-table, is as follows: Timber (in floor), 26,025 feet B. M.; wrought-iron, 495,575 pounds; cast-iron, 122,041 pounds.*

In proportioning the draw, it was supposed to carry the whole dead lcad on the central bearing when swung, and each arm was supposed to carry its share of the dead load, and a moving load of one ton to the foot when closed, no allowance being made for the continuity of the chords. Though this has been the method by which most of the large iron draws lately built have been proportioned, the engineers were convinced that it is a method of computation which gives very erroneous results, showing the central strains, especially in the web, to be much less than they really are, with corresponding excesses in other parts; a set of calculations believed to be based on a more correct hypothesis will be found in a subsequent chapter. The distribution of strain is regulated by the proportion of the total weight thrown upon the end piers, and is therefore largely dependent on the form of latch used. The wedges under the end posts have but a small lifting power, as is fully proved by the action of the draw under a passing load, a heavy freight train, covering one arm only, causing the further end to rise from its bearings of an inch. A set of hydraulic jacks are to be substituted for the wedge plates, the jacks being placed within the hollow end-posts and worked from the turn-table by pumps driven by the steam

*The Plans of the Draw are given on Plate XI.

engine; it is thought that under this arrangement a sufficient lifting power can be obtained to make the proportioning of the draw sufficiently correct to prevent distortion.

The 130 and 176 foot spans, as well as the little iron span, were raised in the spring of 1868. The 130 foot span was the first erected, the trestle used in building Pier No. 1 being made available for one side of the false-works. The two other spans were over dry land at the time of their erection, and ordinary false-works, resting on the ground, were used for their raising. The remainder of the superstructure was not raised till the spring of the following year, when the first span raised was that between Piers 5 and 6, while the sand bar continued dry. The greatest difficulties occurred in the case of the span between Piers 3 and 4, where the strength of the current and depth of the water, especially near Pier No. 3, would have carried away any common false-works in a very few hours. The distance between the caisson around Pier No. 3 and the false-works at No. 4 was divided into four nearly equal spaces. Between the first and second of these spaces, a cluster of eight piles in two rows eight feet apart was driven in thirty-five feet of water, the piles being kept from washing out by guying them with lines as fast as driven; a crib of round timber was then built, enclosing the piles, which, on being sunk by filling it with stones, should at once retard the wash and bind the piles together. A precisely similar arrangement was adopted between the second and third spaces. This work was begun on the 10th of March; on the 14th the weather became very cold, and the ice began to run in large quantities; the numerous obstructions of the false-works impeded the flow of ice, and in the forenoon of the 16th it jammed at the bridge site and the river became closed. The weather had already begun to moderate, and in the afternoon of the same day the ice moved out; it was very weak, but the cakes had packed together, forming large thick fields, which, however, were too soft to bear the weight of a man. The sixteen piles of the two clusters had been driven, a crib had been built about the first cluster, though not sunk, and carpenters were at work upon the second crib, when the ice began to move across the whole width of the river at once; it tore out all of the sixteen piles, taking the cribs with them, and carried along with it the pile-driver, barges and men. The boats moved but slowly, being frequently

retarded by ice jams, and while still opposite the town they were overtaken by the steamboat and secured, having suffered no material damage; but no trace of piles or crib-work remained, and two months later one of the cribs was observed forty miles down the river, with a pile still remaining in it. This gorge was accompanied by a considerable scour, the water at the site of the second set of piles having been deepened about twelve feet.

The piles were at once replaced, and the cribs built, sunk and protected by additional riprap; the piles were then capped and surmounted by trestle piers, which were planked on the sides and provided with timber starlings, eight or ten feet high, as a protection against drift. Between the third and fourth spaces, where the depth of sand was much greater, a single row of five piles was driven, which were braced to the false-works of Pier No. 4, riprapped and surmounted by a trestle bent. Other bents were raised on the caisson surrounding Pier No. 3, and on the false-works at No. 4, and a false pier of timber was erected immediately south of Pier No. 4, the masonry being still unfinished, one side of which rested on the upper section of the caisson, and the other side on the false-works. Eight trussed girders, made of track stringers and rods which had been used at No. 4, were built upon the shore, and raised by a floating derrick into position on the trestles, one being placed under each bridge chord; on these were laid the cross-timbers and other staging required. These false-works proved amply stiff, and when removed after the erection of the span, it was found easier to break the piles off immediately above the cribs than to withdraw them. The false-works between Piers 4 and 5 were built at the same time, resting on piles, and a light track was laid from the north end of the bridge, nearly to Pier No. 3.*

The 198 ft. span was raised as soon as these false-works were ready. As Pier No. 4 was still incomplete, the last panel was not put in, but a bearing was taken on the false pier, one panel short of the end of the truss, the top stood projecting over, the links of the bottom chord being left to hang down. On the completion of the pier the last panel was added, and the bearing was transferred to the masonry. The erection of the long span between Piers 4 and 5 followed, completing the number of fixed spans.

These false-works are shown on Plate VII.

The draw span was raised on false-works extending from the pivot pier to the upper and lower rests. As the small amount of sand above the rock precluded the driving of piles, these works were built on cribs, two of which, loaded with stone, were placed between the pier and each rest. These cribs were originally intended to serve as the foundation of a permanent draw protection; they were built in the winter of 1868-9; were made thirty feet square, and divided by four cross-walls into nine compartments. The deadening effect of the upper rest and pier on the current, had so checked the scour that the cribs did not settle to the rock, and as their bearing was not thought to be firm enough to carry a permanent structure, they were built up above ordinary high-water, and a wooden truss, strong enough to sustain itself if the cribs settled, and which should serve as false-works for raising the draw, was built upon them.*

As soon as Pier No. 2 had been completed, the pivot was placed upon it, and the turn-table put together; the chords were then spread out and riveted, and the bridge trusses made self-sustaining at the earliest possible moment, the whole structure being raised in about six weeks. The cribs settled slightly under the weight of iron, but not enough to give trouble, the subsidence being remedied by additional blocking. Since then the upper cribs have not settled materially, and are probably on their permanent bearing; but the night after the weight of the truss had been taken off the false-works, a rise in the river scoured around the two lower cribs, causing them to settle away from the truss; under the continued scour of the summer flood they continued to settle, tilting from side to side, and finally, when the flood was at its height, they tipped over and rolled away; the false-work truss remains standing, and no harm was done to the works.

On the occasion of the public opening on the 3d of July, the bridge was tested in the presence of a number of engineers invited to examine it, with the following results :