IRONBRIDGES

Iron Bridges. THE exclusive use of iron in the construction of bridges is of modern date, though no other material is so peculiarly adapted to such a purpose; it was not, however, from any want of appreciation of these advantages, so

much as from the great cost, and even impossibility, of obtaining iron in large masses, that its use was so long delayed. It is now most extensively employed in bridge construction, and though, in elegance or durability, iron cannot compete with stone, where the span is moderate, yet there are numberless cases where its adoption has been the means of solving many of the great problems of modern engineering, and its use has more especially become an absolute necessity in Railway Bridge construction, where headway is so frequently of paramount importance,—and where rapidity of execution is often a more necessary consideration than even economy or durability,—while the defective foundations that have so often to be contended with, render the lightness, the independent strength, and pliable character of iron for such structures of the utmost value.

We shall confine our remarks in this article to rigid structures. It is not, therefore, our intention to treat of the Suspension Bridge, which is, moreover, not of such modern date, nor have any important modifications been made in its construction since the first magnificent specimen of its class was erected by Telford over the Menai Straits in 1820. It will, however, come within our scope to describe the attempts that have been lately made to render the suspension bridge sufficiently rigid for railway traffic, and to point out the difficulties of such a problem.

Previous, however, to any description of the various forms of iron bridges now in use, we shall give a brief history of their introduction.

According to Gauthey,¹ the history of iron bridges

¹ "Traité de la Construction des Ponts," par M. Gauthey.

commences in the sixteenth century, when such structures were first proposed in some Italian works. In 1719 the subject was again revived by Desaguliers, but nothing like an attempt at construction appears to have been made till 1755, when an iron bridge was proposed at Lyons, which was to consist of three arches of eighty-two feet span; one of these arches was actually put to-

gether in the builder's workyard, but this project was subsequently abandoned from motives of economy, and a timber bridge was substituted.

The first iron bridge actually erected was the semicircular cast-iron arch across the Severn, near the village of Brosely, at Coalbrookdale, in Shropshire, which was commenced in 1777, and completed in 1779. It must

be remarked that some few years prior to this date, the smelting of iron with coke was first successfully accomplished by Mr Abraham Darby of Coalbrookdale, when the use of cast-iron began at once to supersede that of timber in numerous details of construction. This bridge was designed by a Mr Thomas Farnolls Pritchard, an architect of Shrewsbury, and was erected by Mr Abraham Darby, and his partner Mr Reynolds, the proprietors of the Coalbrookdale Iron Works; the expense of its

construction being defrayed by a joint-stock company. This arch is nearly semicircular, the span being 100 feet 6 inches, and is composed of five ribs, placed 4 feet 10 inches apart; the ribs (see fig. 1) are formed of three concentric arcs of cast-iron, of which the lower one only is entire, and made in two castings united at the crown, the other two being deficient at their summits, in consequence of the superstructure resting on the crown of the lower and entire arc; the outer arcs are $5\backslash frac\{1\}\{4\}$ inches

square, and the lower arc is 8½ inches deep, by 5½ inches thick; they are connected by means of radial bars fastened with bolts and nuts; and, at their springings, they are supported upon cast-iron plates 4 inches thick, bedded upon masonry. The roadway is supported on open cast-iron spandrels, and is formed of a bed of clay mixed with foundry slag laid upon cast-iron plates. The total weight of iron-work is 378½ tons.

If we consider that the manipulation of cast-iron was then completely in its infancy, a bridge of such dimensions was doubtless a bold, as well as an original undertaking, and the efficiency of the details is worthy of the boldness of the conception. It is to be regretted that, from a defect in the abutments, and the error committed in treating the arch as one of equilibrium, the abutments were thrust inwards at the approaches, and the ribs partially fractured.

Shortly after the completion of this arch, some bold propositions for an extension of the principle originated with some French engineers. M. Callipe, in 1779, made a design for a wrought-iron arch of 656 feet span; proposing to form the ribs of plate-iron placed edgewise, the thrust being resisted by wrought-iron tie-bars, connected with the ribs by vertical and inclined bars; and M. Montepetit, in 1782, presented a plan for an arch of

213$\backslash frac\{1\}\{2\}$ feet span over the Seine at Paris, of which the ribs were to consist of two concentric arcs 5 feet 4 inches apart, composed of double plates of wrought-iron, 12$\backslash frac\{1\}\{2\}$ inches deep by $\backslash frac\{1\}\{2\}$ inch thick; the segments were shewn with return ends, fixed together by bolts and nuts. An engraving of this bridge was given in the Encyclopédie Methodique. M. Guyton published, in 1782, some observations upon these designs, proposing to form the ribs of three plates, breaking joint with each other, and united by a species of stirrup at each joint, with distance-pieces between the arcs, and bolts passing through all. M. Guyton, moreover, was of opinion that no difficulty would be found in constructing such arches of double the span of the largest existing stone arches.

In 1783, M. Racle gave a design for an iron bridge of three arches of 84 feet span, to be erected at Lyons on the site of the Pont de la Mulatiere. In this bridge, as well as in other designs by that gentleman, the arcs consisted of panel work, or of cast-iron hollow voussoirs.

M. Aubry also proposed a bridge at Cordon, of six wrought-iron arches of 75 feet span each. The ribs were to be formed of two arcs of different radii united by bars, the roadway being suspended from the arches. He also gave a design, in 1786, for the construction of a wrought-iron arch of 318$\backslash frac\{1\}\{2\}$ feet span,² the ribs being formed of two tiers of arcs, united by radial bars, and the spaces between them filled with diagonal bracing.

Though our neighbours were thus so prolific in design, not one of these various bridges was constructed; a circumstance to be accounted for partly from want of enterprise, and partly from the fact that, with the exception of M. Racle's cast-iron voussoir bridge, they were all formed of wrought-iron, a material which, though more employed in France than cast-iron, was not obtainable in plates or bars of sufficient dimensions for these bold projects; accordingly, after a lapse of some years, the next actual construction of iron bridges took place in England.

In 1795, the old stone bridge which crosses the Severn at Buildwas, about ten miles below Shrewsbury, was washed away by a flood, and Mr Telford being engineer to the county, was called upon to replace it, which he did, in 1796, by a single arch 130 feet span (see fig. 2). The arch, which is very flat, is formed of five cast-iron ribs, having a rise of only 14 feet; there are also two outer ribs with a rise of 30 feet, their springings being lower than those of the inner ribs, and their summits reaching to the top of the iron railing of the superstructure. The platform is 18 feet wide, and formed of cast-iron plates, with deep flanges, which, in themselves, serve as an additional stiffness to the system.

The expansion and contraction of the outer arcs being different from that of the remainder, to which they were dovetailed, was, as might be expected, found rather to derange their action than to add any strength to the structure.

The total weight of iron in the above bridge is about 174 tons, and it was constructed by the Coalbrookdale ironmasters at a cost of £6,034.

We now come to a description of probably one of the boldest examples of arch construction in existence—the bridge over the Wear at Wearmouth, which was completed also in 1796.

About the year 1790, Thomas Payne, the well known political author, proposed to construct cast-iron arches of framed open panels, in the form of voussoirs, and with characteristic energy, he put his views to the

test by constructing an experimental arch of 88 feet 6 inches span, which was made by Messrs Walker of Rotherham, and exhibited in a bowling-green at Paddington; this experiment was completely successful. It was intended to have shipped this arch to America, but Payne not being able to defray the expenses, the manufacturers took it back, and the malleable iron was afterwards worked up in the construction of the Wearmouth Bridge.³

It was in the same year that a committee was appointed for investigating the inconvenient and dangerous state of the ancient ferry in the middle of the harbour at Wearmouth; and it being decided that a bridge should be substituted, arrangements were made for erecting one of stone, although this intention was subsequently abandoned, and at the instigation of Mr Burdon, the Member for the county, who had gone into the subject with the Messrs Walker of Rotherham, it was determined to construct a cast-iron arch. Mr Thomas Wilson was accordingly entrusted with the design, and the ideas of Payne with regard to open voussoirs of cast-iron were adopted.

The history of this extraordinary structure is thus as remarkable as its execution. If we are to consider Payne as its author, his daring in engineering certainly does full justice to the fervour of his political career; for successful as the result has undoubtedly proved, want of experience, and consequent ignorance of the risk, could have alone induced so bold an experiment; and we are led rather to wonder at, than to admire, a structure which, as regards its proportions and the small quantity of material employed in its construction, will probably remain unrivalled.

This bridge (see fig. 3) consists of a single circular arch of 236 feet span, with a rise of 34 feet; the springings commence at 95 feet above the bed of the river, and the whole height above low water is about 100 feet, admitting vessels of from two to three hundred tons burthen to pass under it without striking their masts. The ribs, which are six in number, are placed 6 feet apart, and they are composed of cast-iron panels acting as voussoirs. In elevation these panels form three concentric arcs, distanced by radial bars, each arc being 6 inches broad by 3$\backslash frac\{1\}\{2\}$ inches thick, and the radial bars 1 foot 9 inches long by 2 inches broad. Each separate panel contains two radial bars, and a series of 105 of these panels or voussoirs form a rib, and they are connected by wrought-iron arcs fitting in grooves on their faces, to which they are finally secured by screws. The ribs were united with each other by transverse horizontal struts 5 feet 11 inches long, formed of cast-iron pipes, which are placed at every joint of the voussoirs, with flanges at their extremities, by which they are attached to them.

The spandrels are filled in with cast-iron circles, following the contour of the roadway, and thus diminishing in diameter from the springing to the crown of the rib. The superstructure is a strong frame-work of timber planking, which was first coated with a mixture of tar and chalk, to preserve it from rot, and then covered with a road of marl, limestone, and gravel.

The whole width of the bridge is 32 feet. The abutments, which are almost of solid masonry, are 24 feet thick, 42 feet broad at their base, and 37 feet at top. The south abutment has a foundation upon a solid rock, and rises from about 22 feet above the bed of the river. On the north side, the ground not being so favourable, it was necessary to carry the foundations 10 feet below the bed.

² "Mémoires sur Différentes Questions de la Science des Constructions publiques et économiques, par M. Aubry, 1^re partie."

³ See Reo's Cyclopædia, art. Bridge, sec. Iron Bridges.

The total weight of iron is 260 tons, 46 tons of which are wrought-iron, and 214 cast.

The mode of putting together the ribs was so simple and expeditious, that it was accomplished, and the whole bridge thrown over the river, in ten days; the scaffolding, which was very light, was immediately removed, and the bridge was opened for general use on the 9th of August 1796, the total cost being £27,000.

The construction of the bridge, however, occupied altogether a period of three years, and to complete its singular history, it was sold for £30,000 in a lottery, in October 1816.

Independent of the want of material, the most remarkable deficiencies in this bridge are the want of diagonal bracing, and the bad construction of the spandrels which are filled in with light cast-iron circles.

Soon after the centres were removed, the arch, indeed, was found to have deflected laterally eastward to the extent of 12 or 18 inches, but by means of wedges, tie-bars, and braces, it was partially restored to its original form, and the light cross-bracing, as it at present exists, was inserted. The stability of the bridge has been at all times, however, extremely precarious, and ordinary prudence cannot much longer delay its entire removal.

We must here mention that, in 1794, a small bridge was erected at Laason, in Lower Silesia, by M. le Comte Burghauss, consisting of a cast-iron arch, 43 feet span and 20 feet wide, composed of five ribs, formed of several arcs of different radii united by radial bars; the platform was of cast-iron plates. A similar bridge of 20 feet span was constructed at Berlin, upon the Kupfergraben.

In 1797, Mr John Nash took out a patent for some improvements in cast-iron arches, the ribs being formed of plates. Two bridges upon his system were erected over one of the canals in St. Petersburg. Nash also proposed to form an arch of a series of cast-iron boxes, bolted together, and filled with earth or cement.

About the same period M. Gaston Rosnay obtained a brevet d'invention for some improvements in iron bridges; and suggested the employment of the railing itself, in conjunction with arcs of double or triple wrought-iron plates, for ensuring rigidity. The platform was also made instrumental in strengthening the bridge, and consisted of a double layer of cross planks, which formed an arch resting against the ribs, which were tied together by cross tie-bars.

In the same year, M. Gauthey proposed for the Pont de la Cité at Paris, two outer arches of cast-iron, somewhat similar to the Wearmouth Bridge arches, although not

put together in the same manner, the platform being supported between the ribs. Although the Council of Ponts et Chaussées had approved of this design, it was not carried out, two wooden ribs being substituted for the iron ones, which ultimately failed.

In chronological order, we next come to the remarkable design by Messrs Telford and Douglas, in 1801, for replacing London Bridge by a single cast-iron arch of 600 feet span, with a clear headway of 65 feet above high water. The ribs were to be of cast-iron, in segments as large as possible, and they were to be connected by diagonal cross-bracing, disposed in such a manner that any part of the ribs or braces could be taken out and replaced without injuring the stability of the whole, or stopping the traffic. On plan the superstructure was arranged to spread in width from the centre of the arch to the approaches. The scheme was investigated by a select committee of the House of Commons, the works were put in hand, and the river was contracted to the necessary width. But this design was at length abandoned, owing more immediately to the difficulties, with such a headway, of constructing the approaches, which would have involved the formation of extensive inclined planes from the adjoining streets.

Iron bridges now began to be generally adopted. Mr Wilson constructed a cast-iron arch of a span of 180 feet, at Staines-upon-Thames, in 1802 (see fig. 4). Its radius was 260 feet, its rise being 16 feet. It was in many respects similar to the bridge of Wearmouth, being formed of cast-iron panels 4 feet 10 inches long; the concentric arcs were 6 inches deep by 4½ thick. The ribs were connected by cast-iron rectangular bracing-frames,

and the voussoirs were united by dowels; the wrought-iron arcs, such as were used to connect the voussoirs of Wear-mouth Bridge, being dispensed with. The spandrels are filled with cast-iron circles, and the platform is formed of cast-iron plates stiffened by curved webs, so arranged as to cover two and three ribs alternately. The abutments of this bridge were insufficient to resist the thrust of the arch, and one of them actually moved horizontally on its base without disturbing its joints.

The first iron bridge constructed in France was the Pont du Louvre at Paris, in the year 1803. The plans

were prepared by M. Cessart, and modified by M. Dillon, who had charge of its erection. The castings were made near Touronde, at the iron-works of MM. Baudry and Mercin.

The arches of this bridge, which are nine in number (see fig. 5), and 57 feet span, are composed of five cast-iron ribs placed 8 feet apart. Each rib or arc is 6½ inches deep by 3 inches thick, and is made in two parts connected at the crown.¹

The springings rest on cast-iron plates let into the masonry of the piers, which at top are 6 feet 4 inches thick. The rise of the arcs is 10 feet 8 inches. Other arcs of lesser dimensions, striding over the piers, and resting on each pair of large arcs, are employed to fill in the depth of the spandrels or haunches. This bridge was intended for foot passengers;—its total length between the abutments is 516 feet, and its width between the railings 30 feet. The total weight of iron in the nine arches is about 263 tons. The stone piers, which are eight in number, rest on pile foundations. An interesting experiment was made on one of the ribs by M. Dillon; it was placed upon a substantial timber platform, and weighted with several boxes suspended from the same points of the rib from which the platform itself was to be ultimately suspended, the boxes were gradually loaded until the weight they contained amounted to double that which the rib could possibly have to sustain, the rib was found to be depressed at the summit, and raised at the haunches; but on the removal of the weights the arch

¹ The original design for this bridge consisted of eleven arches of 51½ feet span, and the ribs were somewhat similar in external appearance to those of the modified design by M. Dillon, although differently put together. M. Cessart, who was eighty-two years old when he made this design, had the satisfaction of obtaining the approval of the Council of Ponts et Chaussées; and to show the gratification he experienced by that honour, it may be as well to quote his own words:—"Si le Pont des Artes (now the Pont du Louvre) n'a pas été entièrement exécuté comme je l'ai proposé dès le principe, je m'en tiens pas moins à l'honneur d'avoir conçu la première idée de cet ouvrage et d'en avoir ainsi préparé le succès, à l'âge de quatre-vingt-deux ans, et sous les auspices du Conseil général des Ponts et Chaussées."—Description des Travaux Hydrauliques, de Louis Alexandre de Cessart, tom. i.

Iron Bridges. regained its original figure, which circumstance was strictly in accordance with the theoretical views upon the subject.

We now come to the Pont d'Austerlitz, erected at Paris by M. Lemandé. This light and elegant structure consists of five equal arches (see fig. 6), which are segments of circles 138 feet radius, the span being 106 feet, and the rise 10 feet 8 inches. Each arch is composed of seven ribs placed 6 feet 8 inches apart, each consisting of three concentric arcs, which form panels by their intersection with radial bars. The arcs are $5\backslash frac\{1\}\{2\}$ inches deep by $2\backslash frac\{1\}\{2\}$ inches thick, and the radial bars are $2\backslash frac\{1\}\{2\}$ inches broad; the spandrels are filled up by arcs and radial bars, forming panels similar to those of the voussoirs. The ribs are united by cast-iron cross-frames, formed of bars $2\backslash frac\{1\}\{2\}$ inches square by 6 feet 5 inches long, with a double return at their extremities, by which they are attached to the voussoirs; the depth of the entire arc is 5 feet 3 inches. The platform is of timber,

covered with a thick bed of loamy earth and gravel. The joints of the voussoirs were run in with spelter, and an excess of rise was given to the summit of the arches

of $2\backslash frac\{1\}\{2\}$ inches. The settlement which took place on the striking of the centres varied in different arches from $\backslash frac\{1\}\{4\}$ inch to $\backslash frac\{1\}\{2\}$ inch—but it ultimately increased to $2\backslash frac\{1\}\{2\}$ inches and $2\backslash frac\{3\}\{4\}$ inches. The deflection was partly caused by the fracture of some of the radial bars, particularly those

Iron Bridges. near the abutments, caused by the bridge having been overloaded immediately after its completion. The area themselves remained, however, uninjured, and the radial bars having been repaired by bands of wrought-iron placed over the broken parts, the arcs were rendered secure from further distortion.

This bridge was commenced in 1800, and finished in 1806.

In January 1805, some discussion took place as to the best method of widening the waterway of the Pont de l'Archevêché at Lyons, and it was proposed to replace

the five stone arches of that edifice by three cast-iron arches, of which the centre one would have had a span of 239 feet. (See fig. 7.) The proposed construction was novel; the arcs being formed of cast-iron pipes, two tiers of which were proposed for each rib, connected by radial bars, having their ends flattened and bolted between the flanges of the pipes. A similar arrangement was proposed for the bracing-frames between the ribs. The haunches of the arches were also composed of pipes.

The proposed advantages in the use of pipes arose from the consideration that the same quantity of metal in a hollow section was much more advantageously disposed of than in a solid section, and this first application of that principle to bridge structure is deserving of notice.

Some novelty in construction next occurs in a small wrought-iron arch, constructed in 1808, by William

Bruyér¹ near St. Denis; the span is only 37 feet 5 inches, with a rise of 8 feet 3 inches, and the ribs are composed

of panel voussoirs, but the angles of each panel are con-

¹ M. Bruyér had applied this arrangement with advantage to the construction of wrought-iron sluice gates.

nected in a peculiar manner (see fig. 8); the ends of each radial bar have four quadrant plates forged to them, and the arcs and the diagonal cross-bars, which all meet in the same point, have similar quadrant plates, that halve in thickness with those of the radial bars, and circular plates are made to cover the whole, which are fastened together by four bolts and nuts.

Such a method of construction is extremely costly, and evidently inapplicable to large structures.

A similar bridge was, however, proposed in 1810, by M. Bruyer for the prolongation of the centre portion of the Hotel des Invalides, at Paris; it was to be a single arch of 426½ feet span, the ribs being divided into voussoirs similar to that just described, and tied together by cross-bars and diagonal tie-bars at every interval of the voussoirs. The execution of this structure, which was only meant for foot passengers, had been decided upon, but was afterwards suspended.

Another bridge for the same site was proposed by M. Lamandé, in 1811. This design presented some additional peculiarities, inasmuch as it consisted of a combination of wrought and cast-iron; the bridge was to consist of three arches, but the centre one only was to have been of iron, with a span of 262½ feet. In this instance there was to have been a carriage-way. The arcs of the voussoirs were to be of cast-iron, but the radial bars and the diagonal braces, and in fact, all those pieces which were exposed to tensile strains, were to have been of wrought-iron. Each upright served to connect two adjoining voussoirs, and carried, at its extremities, ears or lugs, forming a kind of fork which clasped the voussoirs; the cast-iron arcs of the voussoirs were further bound together by wrought-iron arcs.

Improvements in the manufacture of iron kept pace with the rapid increase of its application to construction, while the ingenuity of inventors in this, as in all other development of new principles, far outstripped the requirements of practice. Mr Pope proposed to construct bridges of 1000 feet span, on a new principle, named by him the "Flying Lever Pendant Bridge." However impracticable the construction may have

been, the idea is worthy of notice, inasmuch as it involved new and important applications of action and reaction. These bridges were to consist of a combination of struts and ties acting with various leverages—and so arranged that the entire half-bridge was one huge cantilever, with the abutments sufficiently massive to bear all the over-hanging weight and strain of the half-arch. By this means, there was no dependent pressure at the crown or meeting of the two half-arches—each of which was to be built progressively from the abutments to the centre, without the aid of centering or scaffolding; with this arrangement, any portion of the middle could be removed without disturbing the rest.

These bridges were to be constructed of iron or timber.¹ M. Wilbeking, Director-General of Ponts et Chaussées, in Bavaria, published, in 1812, a treatise upon Iron Bridges, in which he proposed to construct arches of iron pipes, fastened together by means of sockets attached to the extremities of the cross-bars into which the ends of the pipes entered; the cross-bars themselves were also formed of cast-iron pipes. M. Wilbeking gave a design for two bridges upon this principle, one of 111½ feet span, by 11 feet 2 inches rise, and the other of 292 feet span by 29 feet rise. The pipes were to be about 8½ inches diameter, by 1½ inch thick, those forming the cross-bars being something less. Each of these bridges was composed of three parallel ribs, with two diagonal or bracing ribs. The spandrels were filled in with radial supports, stiffened by diagonal bars, or by wrought-iron rings. The possibility of keeping such a system in form is extremely problematical.

Messrs Jessup erected two iron bridges of great simplicity at Bristol, the spans being 100 feet, and the rise 15 feet. The arch is composed of six ribs made of two castings united at the crown, and each rib is provided with seven uprights, which sustain the platform of the roadway: the ribs are held together by nine wrought iron cross-bars or ties. Each of these bridges contains 150 tons of grey iron, and was constructed at a cost of £4000.

¹ Mr Pope, in his work upon Bridges, expresses his indignation in strong terms against many scientific gentlemen of that day for their seeming little faith in his design.

For examples in iron bridges, we are much indebted to Mr Telford. A very fine iron bridge (see fig. 9), by that gentleman, spans a natural channel formed by the Spey in the Craigellachie rock. The span is 150 feet, and the arch is composed of four ribs, the rise being 20 feet. Each rib consists of two concentric arcs, forming panels, which are filled in with diagonal bars. The roadway is only 15 feet wide, and is formed of another arc of greater radius, to which is attached the iron railing; the spandrels are filled by diagonal ties, forming coarse trellis-work. The bridge rests upon stone abutments, with circular castellated towers and curved wing walls at each approach. This bridge is undoubtedly open to the objection, that it consists of two very dissimilar arches, which, owing to the variations of temperature, must be subject to very variable strains. The total cost of this bridge, including the blasting of rock on the west side of the river, was £8200.

Another bridge very similar, and of the same span, was erected by Telford over Dornoch Frith, called Bonar Bridge.

Several iron bridges, about the same date, were erected by Mr Rennie—one at Boston, over the Witham, has a span of 86 feet, with a rise of only 5 feet. Owing to the unequal cooling of the ribs in the foundry, or from some other cause, they were fractured in several places when loaded, but notwithstanding these fractures, the arch generally preserved its form.

We now come to the description of the two iron bridges erected over the Thames, in London, viz., Vauxhall Bridge and Southwark Bridge, the latter being in every respect a most remarkable structure, whether as regards its bold proportions or architectural effect.

The bridge at Vauxhall was designed originally by Mr James Walker; but in consequence of some disputes in regard to its construction, four engineers were ultimately engaged in its erection, namely, Mr Ralph Dodd, Sir James Barthram, Mr Rennie, and Mr James Walker. It consists of nine arches, the span being only 78 feet, and the rise 29 feet, the piers are 12 feet thick; the roadway is 36 feet wide between the railings, including the footpaths of 6 feet each. Each arch is composed of ten cast-iron ribs 18½ inches deep, with a double flange 6 inches wide, the thickness of the middle web is 1 inch, and that of the flanges 2 inches. Each rib, with its spandrels, consist of three castings; there are, consequently,

only two joints which divide the arc in three equal segments; the spandrels consist of light vertical panel work, the verticals in section forming a cross. The upper line of the ribs is straight, forming two inclines to suit the slope of the road; the ribs are united by diagonal strutting-frames, of which there are five rows in each arch. The ribs abut upon cast-iron plates, which are embedded in the pilaster of the piers and abutments. The roadway plates are of cast-iron, with flanges on the upper side, by which they are bolted together. Over each of the four faces of the middle piers, there is a cast-iron alcove; the railing is 4 feet 6 inches high, and composed of cast-iron verticle bars. The outline of the superstructure is a curve between the middle piers, with two straight inclines, tangents to it. The total length of the bridge is 809 feet. The first stone was laid on the 9th of May, 1811, and the bridge was opened to the public in July 1816. The iron-work was cast at Butterly, in Derbyshire, and the cost was £300,000 for the whole of the work included in the structure.

The Southwark Bridge (see fig. 10), designed by Mr Rennie, crosses the Thames between London and Blackfriars' bridges, and as an example of arch construction, stands confessedly unrivalled as regards its colossal proportions, its architectural effect, or the general simplicity and massive character of its details. The depth of the river being from 30 to 36 feet at high water, Mr Rennie decided on crossing it with three arches; the centre one having a span of no less than 240 feet, and the external arches 210 feet each; the rise of the centre arch is 24 feet, and of the external arches 21 feet. The piers and abutments are built of Scotch granite, and rest on platforms upon timber piles, protected by timber sheeting. The length of each pier, including the cutwater, is 78 feet, and the width 24 feet; the height from the cornice of the centre arch, to the surface of the water at high spring tide, is 42 feet; the total length of the bridge, with the abutments, is 800 feet, and the width of the roadway between the railing is 42 feet, which includes two footways of 7 feet each. Each arch consists of eight ribs, each rib being formed of thirteen vertical cast-iron segments or voussoirs; they do not abut immediately upon each other, but upon an intervening transverse plate, extending across the eight ribs at the junction of every voussoir, a most efficient bond being thus obtained between the individual ribs: there are fourteen such trans-

verse plates on each arch, including the two at the abutments. The voussoirs are 20 feet long, and 6 feet deep at the crown, and gradually increase to 8 feet deep at the abutments; they are $3\backslash frac\{1\}\{2\}$ inches thick, and have double

returns, 4 inches wide at their abutting edges, by which they are bolted to the transverse plate; fillets are cast upon this plate for the introduction of wedges to insure perfect contact throughout the whole depth of the voussoirs. The spandrels are cast independent of the voussoirs, and attached to them by means of a number of dental projections fitting into corresponding recesses in the voussoirs, both of which are dovetailed; their form is exceedingly simple and efficient; they consist of lozenge-shaped panels, formed by diagonal lines, and intersected at about the middle of the depth by a curved line much flatter than that of the arc of the voussoir. Diagonal bracing-frames are introduced between each consecutive spandril, and the whole system is further secured from lateral motion by means of four diagonal lines of bracing-bars intersecting each other and extending across every arch. The roadway is formed of cast-iron plates 4 feet wide, and alternately 22 and 11 feet long. These plates are strengthened by ribs, and are bolted together by flanges at their edges, they break joint with each other and rest alternately on six and four ribs, and thus contribute materially to the lateral strength of the structure. The parapets and cornice, as well as the balustrade, which is 4 feet 8 inches high, are all of cast-iron.

The iron-work was cast by Messrs Walker & Co., of Rotherham, and the total weight was about 5780 tons; each rib was carefully fitted together at the works before it was shipped for London. The total cost, including the approaches, connecting-avenues, &c., &c., amounted to about £800,000. The work was commenced on the 23d of September 1814, and the bridge was opened in April 1819.

No provision having been made for expansion and contraction, considerable inconvenience arose from its effects, the paving stones of the footpaths having been broken

and disturbed in adapting themselves to the change of form of the arches. The enormous weight of this structure contrasts rather disadvantageously with that of others we have described, but the importance of unusual durability in such circumstances was doubtless duly considered.

A short time after the completion of Southwark Bridge, two cast-iron arches were erected by Telford, viz., an elegant arch across the Severn at Tewkesbury (see fig. 11), of 170 feet span, with a rise of 17 feet only, consisting of six ribs about 3 feet 3 inches deep, the spandrels being filled in with light diagonal work; and an arch, of a span of 180 feet, over the Gloucester and Birmingham Canal at Galton, with a rise of 18 feet; in this case, the ribs consist of diagonal panel work.

It will be observed that all the iron bridges we have described have been arches, and the material used cast-iron. Wrought-iron has been but partially introduced, although its use had been suggested for such structures at an early period; it had, however, been used in the construction of suspension bridges, of which the Menai Bridge, by Telford, was one of the earliest and finest examples. It is evident that in many, if not in all of these structures, the arch has been treated as an arch of equilibrium, and that the two principles on which such construction should be based, were but imperfectly understood. The stability of an arch of masonry depends on its own weight, and the equilibrium of its individual parts, for maintaining its form; such also is the case to a great extent with cast-iron arches of colossal dimensions, such as the Southwark Bridge, where the weight of the rolling load bears only a small proportion to the weight of the structure itself; but the theory of equilibrium, as applied to arches of masonry, is evidently inapplicable when the rolling load bears so large a proportion to the weight of the structure, as is the case in most arches constructed of iron. Another important difference

arises from the fact, that arches of masonry owe much of their stability to the friction and adhesion of the large bearing surfaces of the stones which form the voussoirs, whereas, the bearing surfaces of the segments of cast-iron arches are necessarily limited in dimension, and being in all cases attached by bolts, the arch is rigidly restrained from assuming a position of equilibrium and rest. In the Southwark Bridge, this difficulty was partly met by the introduction of the wedges we have described; and

on the Pont d'Austerlitz, by ultimately closing the joints with spelter run in between the voussoirs. It is evident, on these grounds, that a cast-iron arch cannot be considered as a mere system of voussoirs assuming a position of equilibrium, but rather as a rigid structure, consisting of two inclined struts or beams, abutting against each other horizontally at the summit of the arch, and upon this assumption, we can arrive at the true direction and amount of strain, each semi-arch being treated as a

rigid beam. Such considerations render the calculations of the strains in an iron arch extremely simple.

It will be observed, that in all the arches hitherto described, the thrust is resisted directly by the abutments, as in arches of masonry, and not by the tie-rods, as in the ordinary "bowstring girder," they are all therefore true arches, their claim to originality arising from the introduction of iron in lieu of masonry, and from the ingenious and skilful arrangement of detail displayed by the engineers.

We have now arrived at an epoch in civil engineering, which at once enlarged tenfold its sphere of action, and rendered impossible all reference to experience or precedent; and the arch and the beam, as well as every other established principle of construction, underwent, with miraculous rapidity, entire modification, and their application became wonderfully extended.

We allude, of course, to the introduction of railways, in which the application of iron takes an entirely new direction. The success that had attended the use of tramways in some collieries, led, in 1823, to the construction of the Stockton and Darlington line. The Liverpool and Manchester line, however, which was immediately afterwards commenced, and where the locomotive engine

was first successfully applied, ranks as the true prototype of our present system.

Hitherto, bridges had been applied generally to high roads where inclined approaches were of small importance, and in determining the rise of his arch, the engineer selected any headway he thought proper, while every other consideration was similarly made subsidiary to the problem of constructing the bridge itself, and the completion of a single large bridge was an epoch in engineering history. On the introduction of railways, hundreds of roads, rivers, and valleys had at once to be spanned with level roads. Time was as important an element as economy or durability in the erection of these structures, while every conceivable difficulty arose from their limited headway, their bad foundations, their oblique directions, or their gigantic dimensions. Navigable waters, as well as crowded thoroughfares, had now to be crossed without interference with existing traffic, and the ponderous locomotive dashed over these new and hastily constructed works, instead of the quiet team. The arch was evidently at once inapplicable to the bulk of such requirements; new principles of construction became imperative, and the beam, with all its numerous modifications, at once

superseded the iron arch as completely as the locomotive did the stage coach. The earliest simple iron beams of which we possess any account, are those used by Mr Telford in building a cotton-mill in Salford, in the year 1800 as recorded in "Tredgold's Essay on Cast Iron," by Hodgkinson, part ii. But the application of simple cast-iron beams to the construction of bridges originated with the late George Stephenson, who employed them on the Liverpool and Manchester Railway.

It will now be more convenient to discontinue our chronological history, and to avoid repetition, we shall, in the next chapter, give a classification of the various forms of girders now in use on railways, describe their general principles of construction, and proceed with our descriptions of some of the most remarkable examples. Under their respective heads, we here give, in a tabular form, some of the dimensions and other characteristics of the works hitherto described.

Tabular List of the Bridges constructed as described in the First Chapter.

WE have, in the last chapter, traced the history of iron bridges of rigid form down to the re-introduction of the primitive straight beam or girder—apparently no doubt a retrograde step, when compared with the elaborate and elegant structures we have been considering, and on which so much scientific investigation and mechanical skill had been bestowed. We shall find, however, that the beam rapidly outgrew its original simple form and dimensions, and it is now scarcely to be recognised as the parent of those magnificent structures (far exceeding in dimensions the largest arches), which have become our most prominent monuments of engineering enterprise and skill. We cannot fail to be struck with the reverse order of progress which obtains in architectural history, where we find the beam or lintel characteristic of all the early temples, the arch and abutment characterising the progress of scientific construction.

The word beam, or girder, has thus entirely changed its original signification, and some notice of its present extended meaning is requisite. If we consider the various means employed for crossing space, we find, first, that the weight of the structure and its load is transferred to the bearing points on each side of the space to be crossed, always exerting there a vertical force or pressure corresponding with the weight; and secondly, that these vertical forces or pressures are invariably resolved at the centre of the span into direct horizontal forces.

Now, in order that vertical forces may be transformed into horizontal strains, and equilibrium maintained, some fulcra or resistances must be interposed, and equivalent horizontal forces of an opposite character be exerted;

and these fulcra and forces may either exist within the structure itself, where they are termed the "strains on a beam subjected to transverse pressure," or may be external, in the form of corresponding horizontal pressure upon the points of support, in which case the pressure is called "the thrust of the arch;" the oblique direction of the thrust of the arch is thus the resultant of the direct vertical weight of the mass itself, and the direct horizontal reaction of the thrust at the centre.

We have thus two distinct classes of bridges.—First, those in which the horizontal strains are counterbalanced by corresponding horizontal resistances at the abutments, the strain on them being consequently oblique, tending either to draw them together, or thrust them asunder. Suspension bridges and arches are evidently of this class, the simplest type of which would be two abutting struts as in the figure—

Secondly, those bridges in which the pressure upon the abutments is purely vertical, the horizontal strains being met by corresponding horizontal strains within the structure itself, the simplest type of such structures would

Iron Bridges. be two abutting struts with a connecting tie-chain, as in the figure—

It is to every construction of this latter class that we propose to apply the term beam or girder.

All bridge structures will be thus included under these two types. The various forms of beam or girder now in use may, however, be again further subdivided, and we shall adopt the following classification:—

(1.) Iron Arches.

It would be tedious to devote much more space to a description of those structures, after the numerous examples cited, more especially as, in their application to railway practice, no important novelties were introduced.

The true theory of iron arches has been partially alluded to, and it has been shewn that all such structures should be considered as consisting of two inclined struts—it is evident, however, that these struts being curved in the form of the arch, the load will not have an uniform effect upon them—but this adds little difficulty to the problem. The horizontal forces we have shewn to be equal at the springing and centre of the arch; and the vertical pressure which is at each springing equal to half the weight of the centre structure, is, at any other point, equal to half the weight included between that point and a corresponding one on the other side of the centre.

The laws which govern the horizontal and vertical forces being thus known, it is only necessary to calculate their amount and their resultants to shew the magnitude of the force in the direction of the curve at any point. The joints at each point should naturally be at right angles to the direction of these resultants, and the depths of the voussoirs should correspond with the amount of pressures.

The horizontal force, which is everywhere constant, may be found upon very simple considerations. It depends, first, upon the depth of the arch, or rise, as it is termed. Secondly, upon the horizontal distance of the centre of gravity of the half-loaded arch from its springing; and thirdly, upon the amount of load.

In considering the strains caused by the weight of the arch itself, exclusive of the load, it may be well to observe that, supposing the arch to be uniform in section, its weight per unit of horizontal dimension will not be uniform; and therefore, the horizontal distance of the centre of gravity of the half-arch from the abutment must be the element of length in the calculation, and this

horizontal distance will vary, of course, according to the figure of the arch. Similar reasoning must be observed with regard to spandril arches, or those with a horizontal top chord. From the preceding remarks, it is evident that when the weight of the arch itself is considerable as compared with the load, it will be necessary, as we at first observed, to take the horizontal distance of the centre of gravity of the loaded half-arch, which will be at a point between that of the arch itself and that of the load, which, if uniform, will be at the quarter span.

The following simple proportion will therefore give at once the horizontal thrust:—

As the depth of the arch is to the horizontal distance of the centre of gravity of one half of it from the springing, so is the weight of half the arch to the horizontal thrust throughout its length.

Let A B be such an arch, the points A and B being taken at the centre of the springing. Let the rise at the centre, measured from A B to the centre of the middle voussoir, be 10 feet, and let the whole weight of the structure and its load be 100 tons; the girder itself forming but an insignificant portion of that weight, the centre of gravity may be taken at the quarter span, or 40 feet horizontally from the point A.

Then as $10\; :\; 40\; =\; 50\; :\; 200$ { Tons. Tens. }
{ The horizontal strain and
the centre and abutments.

The same reasoning applies to all the description of beam that we have tabulated. In the "Bowstring," the horizontal thrust of the arch or bow is resisted by the tensile strength of the tie-rod or chain; and in the plain girder, by the tensile resistance of the bottom web. To determine the horizontal strains in all these various forms of beams a very simple rule is given in "Britannia and Conway Bridges," page 195, which is—

where S is the horizontal strain in the middle of the top or bottom member of the beam, W the total uniform load including that of the beam L the span, and d the depth (or in an arch the rise). Applying this rule to the example just given, we shall have—

which agrees with the method before used.

We will call attention to another most useful and practical formula, explained also at page 196 of "Britannia and Conway Bridges," and which is of extremely general application, namely,—that in all arches, suspension bridges, tubes, flanged girders, or bowstring girders, and, in fact, every similar structure with top and bottom members, whenever the depth is $\backslash frac\{1\}\{4\}$th of the span, and the load, including the weight of the structure, is equally distributed, the horizontal strain in the middle of both top and bottom members of the system is exactly equal to twice the weight of the system with its distributed load.

Again referring to the example which is given in the proportions mentioned:—

It must be remarked that, in solid beams, trellis beams, or any of those which have their top and bottom webs attached by intervening struts or continuous plates, the horizontal strains in such top and bottom webs are not, as in the tie of the "bowstring," everywhere equal; but they diminish as they approach the bearings, where they become zero. The rules we have given, when applied to these beams, only determine the horizontal strains in the middle.

Towards the end of this chapter, when describing beams of this class, we shall treat more fully of the theory of the various strains to which they are subjected. As regards the arch, from what we have shown, the reader will meet with no difficulty in finding the horizontal thrust, with the centre of gravity of the semi-arch, in any position; and, as we propose to confine ourselves to the principles of our subject, we shall leave to the reader all mathematical application, merely observing that a simple practical way of discussing the arch will be found to be its imaginary transformation into a beam, and the application of similar reasoning to its investigation.

The requirements of railway practice have by no means been favourable for the introduction of the iron arch the height required, from the circumstance that the roadway must be over the top, the practical difficulty of meeting the thrust, and the necessity of a perfect stability in its foundations, have all been drawbacks to its use, although some of our most elegant and efficient railway bridges are cast-iron arches.

The earliest examples we have of such an application of cast-iron arches are in a series of three bridges for crossing the London and Birmingham Railway over the Grand Junction Canal at Blisworth, Boxmoor, and Nash-mill.

That at Blisworth is of 50 feet clear span, and consists of six cast-iron arched ribs, having a rise of 8 feet; the four inner ones are arranged in two pairs, one under each line of rails; the two ribs composing each pair being 4 feet 11 inches from centre to centre, and a 6-feet space between each pair; the two single outer ribs are again 6 feet from those in pairs; making a total width from centre to centre of the outer ribs of 27 feet 10 inches. The depth of the ribs is 2 feet 3 inches at the springing, diminishing to 2 feet at the crown; their thickness is 2 inches, with a projecting flange 6 inches wide at the top and bottom; making a total sectional area of 51 square inches in each rib at the crown. They rest on cast-iron skew back plates, and these again on blocks of stone let into the brick piers. The ribs are each made in three equal segments bolted together, and are connected by a system of trussing which we shall hereafter describe. The haunches are filled in with three separate castings, one over each spandril, and one as a saddle over the crown. The pattern consists of bars of a cruciform section, crossing each other diagonally, and forming diamond-shaped panels, decreasing in size towards the crown, whose upper apices are connected together by a rib or top table, and their lower ones connected to the main rib by being keyed in between projections upon it. Upon the top tables just mentioned, and firmly bolted to them, are placed strong cast-iron plates 3 feet wide, $\backslash frac\{3\}\{4\}$th of an inch thick, with flanges $4\backslash frac\{1\}\{2\}$ inches deep all round, and diagonal flanges from corner to corner. These answer the double purpose of steadying and bracing together the spandrils, and also of carrying the ballast, which, however, is not used for bedding sleepers, the rails being carried in chairs resting on longitudinal balks of timber, and bolted down through them to the top table of the spandrils.

The peculiar feature of the bridge is in the system of trussing employed to connect the main ribs. At equal distances along the curved rib there are cast-iron struts fur-

nished with skew ends, with bevelled edges, for the purpose of keying them in between projections cast on the main ribs; these struts are 12 inches deep, 2 inches thick, and all radiate towards a line joining the centres of curvature of all the arched ribs. The skew end of one strut is placed opposite that of the strut on the other side of the rib, so that they form a zig-zag line, the general direction of which is parallel to the abutments; between these struts are placed distance pieces of a cruciform section, with broad ends with bevelled edges, and keyed in between projections on the struts, in the same manner as the latter are fixed to the main ribs. The skew-back plates before mentioned extend the whole length of the abutment, and are of an irregular shape, so as both to fit the springing of the arch, and also to radiate in the same manner as the struts above mentioned, forming, indeed, the last of these struts on each side.

It will be observed that the whole of the bearing portion of this bridge is put together without any bolts (with the exception of those at the junctions in the main ribs, and those fastening the platform plates to the top tables of the spandrils), every junction being made with keys in the manner just described, so as to render motion in the joints almost impossible, and to assimilate the system, as it were, to one entire piece.

The bridge, indeed, though not perhaps remarkable for its great span, is one that justly deserves notice for the extreme care bestowed by the designer on the minutiae of all its parts, and the great rigidity given to it by the system of trussing so well adapted to the purpose.

The bridge over the canal at Boxmoor is of much the same description; the arrangement of the ribs is precisely the same as of those at Blisworth, but the span and rise are greater; the former is 66 feet, and the latter 11 feet 9 inches, the depth being 2 feet 9 inches at the springing, diminishing to 2 feet at the crown; the spandrils are similar to those at Blisworth, and also the top plates, except that they are lozenge-shaped instead of square. The trussing of the main ribs is different, the system of keying not being so entirely adhered to. It consists of cast-iron struts of a cruciform section keyed in between the main ribs in lines perpendicular to the abutments, and of cast-iron pipe struts, through the entire length of which wrought-iron tie-rods pass, and which extend from side to side of the bridge parallel to the abutments. These also are keyed in between the main ribs, but are not in any way connected with the struts described as being placed at right angles to the abutments.

The bridge at Nash-mill is precisely similar to the one at Boxmoor.

The outer ribs of these bridges are surmounted by light cast-iron railing which extends some distance each way past the arch to the top of the slopes on either side, and gives the bridge a neat and pleasing appearance.

There is also an elegant bridge on this principle, erected on the Midland Counties Railway over the River Trent, at its confluence with the Soar, near Sawley. (See fig. 15.) It consists of three segmental arches of 100 feet span, the radius being 130 feet. They are similar in principle to the arches of Vauxhall Bridge, but possess considerable architectural beauty, the spandrils of the ribs being formed with vertical panels in the Tudor style. Each arch consists of six ribs, strutted and tied by bracing-panels and rods; the ribs are double-flanged, and the rise is $\backslash frac\{1\}\{2\}$th of the span; the depth of the rib is 3 feet at the springing, and 2 feet 6 inches at the crown; they are each in three castings, and the work is admirably executed. The platform is of timber; the width between the balustrades being 27 feet. This bridge was completed by the Butterly Company in 1839; its situation amidst fine scenery adds much to its successful artistic effect.

Numerous other examples might be given on all our

railways, while the advantages which such structures possess, as regards durability and beauty, would warrant their

much more frequent use, where circumstances would allow of their adoption.

Among the boldest structures of this class which have been contemplated, the arches designed to cross the Menai Straits, on the site of the present tubular bridge, are perhaps the most remarkable. These arches, two in number, were of great architectural beauty, the span being 350 feet, with a rise of 50 feet. They are engraved and described in Britannia and Conway Bridges, page 17, and plate 32. The process proposed for their erection was equally remarkable. It was necessary to dispense with all centering, which was to be effected by resisting the horizontal thrust of each successive voussoir by tie-rods passing over the pier, and attached to a corresponding voussoir on the opposite side. Each semi-arch would thus form a bracket tied back by an opposite semi-arch, and the horizontal thrust being thus entirely destroyed, the semi-arches would exert no pressure on each other when they met over the centre of the span, while the whole of the rods employed would only equal in their combined section the sectional area of a chain of sufficient strength to resist the horizontal thrust of the complete arch; the consideration of the strain on these individual rods affords an instructive illustration of the actual strains on the voussoirs of an arch.

Before leaving this part of the subject, we may allude to the introduction of deep iron arches in the construction of roofs, of which an interesting example exists at the Great Western Railway Station at Paddington, and in the roof of the transept of the Crystal Palace. The investigation of the strains in the arch in such cases is extremely difficult. The unequal distribution of the load, the great degree of curvature, and consequent great vertical depth from the springing, together with the attachment of the arch to its supports—all complicate the problem.

These curved ribs, however, are not true arches, since they may be considered as nearly free from thrust.

(2.) Suspension Bridges.

Under this head, we only intend, as before explained, to notice the attempts that have been made to render such structures sufficiently rigid for railway travelling. We may, however, observe, that similar reasoning to that which we have applied to the arch may be applied to suspension-

bridges,—i.e., we may consider them as girders, by imagining the existence of a horizontal strut to resist the tension of the chains.

In comparing a suspension-bridge, or an arch with a girder of the same span, depth, and strength, it is evident that while the suspension-bridge will correspond with the lower member of the girder, the arch will correspond with the upper member, therefore there must be a theoretical advantage possessed by the suspension-bridge and arch in weight over the girder; but the arch is a structure which requires so many additions to preserve it from change of shape, that it cannot practically be taken as merely suffering compression. With the suspension-bridge the case is entirely different; the chain being in a state of equilibrium preserves its form, independent of all extraneous assistance, possessing unaided all the qualifications for resisting the strains of a horizontal load. The suspension-bridge is thus, undoubtedly, the lightest of bridge structures; and the only drawback to its application to railway purposes is its flexibility, which renders it unfit for the support of a rigid roadway. To overcome this difficulty, and combine the rigidity of the girder with the advantages of the chain, has been a problem which has naturally occupied the attention of engineers.

The construction of a suspension-chain for the erection of the tubes intended for the Britannia Bridge, for a long time occupied attention; but this was finally abandoned, owing to the difficulty arising from the expansion and contraction of the chain, which it was proved in a span of 460 feet, with versed sine of 40 feet, would amount to 4 inches. Such a degree of extension would increase the versed sine, and lower the centre portion of the tube as much as 9½ inches. Unless, therefore, the roadway be pliable enough to follow this descent and rise of the chain, it is evident that the chain would be at times either sustaining the whole weight or none at all. The trussing on any rigid roadway suspended from a chain must, therefore, be sufficiently pliable to yield to this motion of the chain, and to be uninjured by its constant action, consequently it must be subject to considerable deflection on the passage of a heavy load. The requisite bracing for securing even this partial rigidity would add, moreover, great weight and cost to the structure, so that, in fact, the economy of the chain would disappear; and, where circumstances will permit, it is far preferable to

convert the chain into the lower member of a rigid beam, as in the Chepstow and Saltash bridges by Mr Brunel, which will be described under the head of Trussed Girders.

In the crossing of the River Niagara over the Niagara Falls, where scaffolding or the floating of any kind of structure to its place were equally impracticable, a suspension bridge was the only possible expedient; and the skilful and successful manner in which the difficulties of such a problem have been met deservedly characterize this work as one of the most remarkable that claims our attention.

The Niagara bridge which has been lately completed by Mr Roebling is indeed the only successful railway suspension bridge of large span. It crosses the Niagara River at a height of 245 feet above the water by a single span of 821 feet 4 inches, and forms the connecting link between the American States and Canada.

The superstructure may be best described as a hollow rectangular box, 18 feet deep and 24 feet wide, on the top of which the railway is laid, while the bottom, which is 25 feet wide, forms the roadway for public traffic—both these floors are constructed of timber beams; and each connecting side consists of a row of double posts or uprights of timber, each pair being 5 feet apart; between them wrought-iron diagonal bars are made to pass, extending each way to the fourth pair of posts at an angle of 45 degrees. The upper or railway floor is suspended from two wire cables at intervals of 5 feet, and the lower floor is suspended at similar intervals from two other wire cables which have a deflection of 10 feet more than the upper ones; these cables, four in number, are each 10 inches in diameter, and composed of seven strands, each containing 520 wires, making a total of 3640 wires. One strand forms the axis round which the other six are twisted; a section of the cable is shown in the accompanying figure. Sixty wires are equal to 1 square inch of solid section, therefore the total area of each cable is 60.4 square inches; or the total sectional area of iron supporting the structure is 241.6 square inches.

Each cable rests upon a separate saddle, there being two on the top of each of the four towers. The saddles are placed on ten cast-iron rollers, 5 inches diameter and 25½ inches long, which bear upon cast-iron plates 8 feet square and 2½ inches thick, strengthened by three parallel flanges which form two compartments for the reception of the saddles.

The ends of the cables are attached to cast-iron shoes, in each of which is inserted a wrought-iron pin which forms the connection with the anchor chains. These anchor chains are each imbedded in a solid shaft of masonry 7 feet by 3 feet, enlarged at the bottom to form a chamber 8 feet square cut in the rock. The shafts are sunk to a depth of 25 feet on the New York side, and 35 feet on the Canada side.

Each anchor chain is composed of nine links, the eight lower links being 7 feet long, and the ninth or uppermost 10 feet long. The lowest link consists of seven wrought-iron bars, 7 inches by 1¼ inches each, and amounting collectively to an area of 69 square inches. They are secured to a cast-iron anchor plate, by a pin 3½ inches diameter. From the fourth link the chain curves, and the section is gradually increased to an area of 93 square inches. There are two towers at each end of the bridge, based upon a mass of masonry 60 feet by 20 feet, which is pierced by an arch 19 feet wide, forming the entrance to the lower roadway. The towers are 60 feet high, 15 feet square at the base, and 8 feet square at the top.

Above the floors are 64 diagonal stays, extending from the saddles to the suspenders, amongst which they are equally distributed; they are formed of wire-rope 1½ inches dia-

meter. There are also 56 stays attached at their upper extremities to the soffit of the bridge, and at their other ends well anchored to the rocks below. The superstructure is thus tied down as well as suspended, and all undulations directly resisted.

The weight of the bridge is estimated as follows:—

The total weight being..... 1658,722 lbs. = 745 tons.

The bridge was commenced in September 1852, and opened for traffic in March 1855. The total cost was £80,000.

It may scarcely be necessary to observe, that were a span of this magnitude required to cross a navigable channel, it would be impossible to adopt the system of stays we have described, and which are so essential in rendering the bridge sufficiently rigid for railway traffic. The engineer has here most successfully and judiciously availed himself of this and every other advantage which the peculiar site afforded, though, under ordinary circumstances, suspension-bridges are not applicable to such traffic. Even in the bridge we are describing, it is found necessary to limit the speed of the trains to three miles an hour.

(3.) Simple Beams.

Under this head we shall include not only rectangular prismatic beams, which are but little used in bridge construction, but more especially that particular form of beam with top and bottom flanges, to which the term flanged girder is generally applied, the vertical web of such girders consisting either of open work or solid plate.

The investigation of the strength of a prismatic beam to resist the effect of a horizontal load, has engaged the attention of mathematicians from the time of Galileo, who considered that a beam loaded horizontally was only subjected to one strain, and consequently only exerted one resistance, viz., that of tension, or, in other words, a beam under pressure would turn on a fulcrum close to that pressure; but by the able experiments and investigations of our more modern philosophers this theory was shown to be fallacious, and we are greatly indebted to the mathematical treatment of the subject by Marriotte, Coulomb, Young, Barlow, Tredgold, and others, without which, and the more recent investigations by Mr Eaton Hodgkinson, we should never have had those magnificent triumphs of art and science which span the widest rivers, and which form the subject of our article. Yet the problem is by no means completely solved, even at the present day, for there still exists a great discrepancy with respect to the calculated and actual strength of a beam.

When, supported at each end, a beam is loaded horizontally, it is found by experiment that its upper surface suffers compression, and its lower surface suffers extension; such compression and extension evidently originating from the vertical support of the beam at each end. Each half of the beam may, therefore, be regarded as a lever in inducing these forces. Now, the beam is imagined to consist of a succession of indefinitely thin horizontal layers, and the compression and extension of these layers is found to be greatest near the upper and lower surfaces of the beam, and to diminish in some ratio towards the centre; it is, therefore, presumed that some layer must exist within the beam in which neither compression nor extension takes place. The position of this layer is called the neutral axis. Now, presuming such to be the case, the extension or com-

Iron Bridges. pressure of a layer must be as its distance from the neutral axis, and the force will be as the extension, if we admit the truth of Hook's principle, Ut tensio sic vis; also the force of a layer must be as its area, consequently the power of any layer to resist a horizontal load is as its force, its area, and the square of its distance from the neutral axis, and by summing up the moments of the indefinitely thin layers, we arrive at the well-known formula—

which also applies to a solid prismatic beam loaded as in the figure:—

Fig. 17.

where $W$ is the load, $l$ is the horizontal length from the fulcrum, $d$ is the depth of the beam, $b$ its breadth, and $C$ is the constant derived from experiment. In the figure, however, the top is in tension and the bottom is suffering compression; in fact, the support in the middle may be considered as a weight acting upwards while the weights $WW$ have the same effect as supports acting downwards.

It is remarkable that the importance of depth in an ordinary beam should so long have escaped notice. We find the floors of old houses invariably supported on massive square beams, with no regard to economy of material, or its proper disposal to resist strain. The use, however, of a more expensive material, such as cast iron, called attention to this important subject, and the solid beam merged slowly into the flanged girder; for it was an easy step, when the effect of depth in a beam was carefully considered, to perceive the value of the addition of a top and bottom flange, thus introducing the material where it was evidently acting to the greatest advantage.

The earliest flanged girders used in bridge construction were formed with their flanges of equal strength, and their vertical ribs were much more massive than requisite for the office which they had to perform, namely, that of simply uniting the flanges. The strength of such beams was calculated by peculiar formulae highly mathematical and complicated, inasmuch as due account was taken of the strength of the middle web to resist horizontal strains. The mode of calculation generally adopted was to consider the section as a rectangle of the entire depth into the width of the flanges, and then to deduct the value of the side portions, the remainder being the strength of the flanged beam.

The rule deduced from the above reasoning was as follows:—

in which $W$ is the load to be sustained, $b$ the entire breadth, $d$ the entire depth, $b\_1$ the breadth, exclusive of the thickness of the web, $d\_1$ the depth between the flanges, $l$ the length, and $C$ the constant quantity, as before.

Upon the preceding mode of reasoning, it is assumed that when the outer layer is strained beyond its ultimate strength, the fracture of the entire section must ensue.

It seems reasonable to suppose that, with the above data,

Iron Bridges. it would be only necessary to obtain by experiment the ultimate force of a unit of section to resist a direct strain (represented by $C$ in the foregoing rules), to enable us to calculate the strength of a prismatic beam. Yet, as we have before observed, such a calculation gives a result widely different to that of practice, the actual strength of the beam being invariably much greater than is determined by the calculation. Some assumption that has been made must therefore be erroneous, and the difficulty of the problem is in accounting for this anomaly. By some it has been assumed that the position of the neutral axis is not situated in the centre of gravity of the section of a beam, but is much nearer the compressed side. This, to a certain extent, is going back to the exploded Galilean theory, yet it is a singular fact that experiments agreed very nearly with this view. (See Experimental Researches on Cast Iron, by Mr Eaton Hodgkinson.) Subsequently, in a paper by Mr W. H. Barlow, read before the Royal Society, on a New Element of Strength in Beams subjected to transverse pressure, and called the Resistance of Flexure, an account of some valuable and interesting experiments was given, which were conducted with the greatest nicety, and which seem to prove that the axis of motion is in the centre of a uniform section; therefore the discrepancy before named has yet to be accounted for, and it seems evident that a great error is committed in considering the existence of a neutral axis as it is called, which suffers neither extension nor compression, the fact being that if it is inactive it is only in a horizontal direction, for there is no part of a beam sustaining a load that is not strained in some direction; and wherever a particle of which it is composed can be shown to be free from strain in one direction, in another direction that particle bears its full share of strain. It appears evident that the vertical reaction of the supports is not entirely met by the simple horizontal resistances of an indefinite number of independent layers. The layers should not be considered as independent like the leaves of a book, since they form altogether a solid mass; and we cannot imagine a top layer to be compressed say $\backslash frac\{1\}\{100\}\; \backslash delta$ of its length, and the next below it $\backslash frac\{1\}\{100\}\; \backslash delta$ without the top layer receiving considerable assistance from its less strained neighbour to which it is so intimately united, and which, in its turn, receives successively similar assistance from every other layer between itself and the neutral axis. The external layer is not therefore subject to the strain which is assumed on account of this assistance which it receives from subsequent layers, and fracture does not take place so soon as such a theory foretells.

The mode alluded to of dealing with the forces brought into action on a beam, when subject to transverse strain, must therefore lead to a misapprehension regarding the true directions of these forces; for instance, the neutral axis is supposed to be a point situated between the portion of the beam undergoing compression and that undergoing extension, and consequently neither subject to extension nor compression, it being regarded as absolutely inactive. Now, this would only be true if the forces called into play were truly horizontal, which in fact is what the present theory of beams assumes.

If, however, we consider the relative motions which take place amongst the particles comprising the upper portions of a beam when undergoing compression, and those undergoing extension in the lower portion, it will at once be perceived that an infinite number of diagonal lines, passing through the region near what is called the neutral axis, may be drawn exhibiting the direction of a variety of forces, both compressive and tensile, and which are not represented by the horizontal lines alone taken into account by the above theory. It is conceived that this mode of looking at the subject shows more clearly than any other that no actual neutral axis can exist; that is, no portion of a beam under

strain can be in a quiescent or passive state. To illustrate this more clearly, let us refer to the figure which represents the centre portion of a beam, the vertical line $bd$ being in fact the centre line.

Let us suppose the beam to be loaded at the top point $b$, then when deflection takes place, the points $a$ and $e$ will recede from each other, and at equal distances from the point $d$, but the other extremities of these diagonal lines will be fixed in the point $b$ by equal forces, consequently there will be a direct strain down these lines of tension. Again, the diagonals $de$ and $df$ suffer compression, being shortened by the approach of the points $e$ and $f$.

Innumerable other diagonal lines in all possible angles might be drawn from these points $b$ and $d$, in which the forces of extension and compression might be traced crossing each other in all directions; therefore, as we have before observed, when a beam is subjected to such a strain as is caused by a horizontal load, and by which pressure naturally takes place, there is no part of such a beam in a quiescent state. Now, respecting the neutral axis referred even to horizontal strains, it is assumed to be free from such strains by the following geometrical reasoning,—viz., if a rectangular parallelogram be bent into a curve, the outer line being extended and the inner line shortened, there must be some line between the two which will neither be extended nor shortened; but it does not follow from this geometrical fact that the line is free from strain because it is unaltered in length. A beam is curved by the effect of strain throughout, and every line in it, excepting the vertical lines, is curved, including, of course, that called the neutral axis; and the fact of that line being curved shows evidently a molecular disturbance which is certainly produced by force, and therefore it is strained.

The other assumption in the above theory, which renders it dependent upon the kind of material of which the beam is composed,—viz., that the reaction of a compressed or extended fibre is proportionate to the amount of compression or extension,—has within moderate limits been proved to be correct as regards wrought iron and cast iron, the amount of the extension or compression caused by 1 ton in a bar 1 inch square, being respectively $\backslash frac\{1\}\{25000\}$th of its length, and $\backslash frac\{1\}\{25000\}$th of its length, or twice as great for the former as for the latter; similarly, if the pressure or tension is twice or three times as great, the alteration in length will also become twice or three times as great, but as we approach the ultimate strength of the iron this law entirely ceases, the compression or extension then increasing in a much higher ratio than the force applied. The assumption, ut tensio sic rit, on which this theory of a beam is founded, only therefore holds good within moderate limits; and this variation, combined with the erroneous assumption of the action of the fibres as being independent layers, before alluded to, gives rise to the anomalies which exist between theory and experiment, and renders the solution of the problem extremely difficult.

To Mr Eaton Hodgkinson the credit is more especially due of pointing out the proper relative proportions of the top and bottom flanges; but the section which these proportions rendered it necessary to adopt, caused the true valuation of the entire section to be so tedious, and even difficult, in the summation of the moments on each side of the neutral axis,—which in this case was not in the mid section,—that that gentleman thought it far better to consider the vertical rib as useless to resist horizontal strains, giving it only sufficient strength to resist lateral strains, to connect the top and bottom flanges, and transmit the strains between them. Therefore, without any complexity of formulae, the

strength of the beam was considered to exist in its flanges, and to vary simply as their distance apart, or as the depth of the beam.

The investigations made by Mr Hodgkinson led directly or indirectly to a great part of the information we possess respecting the properties of wrought and cast iron. He found the ultimate resistance of cast iron to be as follows, viz., the extension 6.5 tons per square inch, and in compression 40 tons per square inch. He determined, moreover, the variation in length within certain limits of a cast-iron bar 1 inch square subjected to strain to be about $\backslash frac\{1\}\{25000\}$th of its length per ton. Similarly with ordinary wrought iron, the ultimate resistance was determined to be 18.6 tons per square inch in extension, and 16 tons in compression, and the elongation of an inch bar $\backslash frac\{1\}\{25000\}$th of its length per ton. He applied these practical results in determining the flexure and strength of girders and pillars, and corroborated the accuracy of his deductions by numberless experiments rivaling in practical value his earlier investigations.

If we neglect the vertical rib of a girder, a very simple view may be taken of its action in resisting a horizontal load, and the result will agree exactly with what we have given with respect to arched beams. In fact, the rule becomes universally applicable to beams generally, and may be briefly explained as follows:—

Let ABCD be a girder, loaded at W, and supported on the points A and B. The forces are transmitted along the diagonal lines AW and BW. If we now construct the parallelogram of forces on the diagonals, we shall have $Wg$ representing the vertical force, or half the weight transmitted to the bearings A and B, and $pg$ will represent the horizontal strain at the centre of the beam, or, taking the whole depth, $Wn=d$, and $An=\backslash frac\{1\}\{2\}$ the length, we shall have—

which is the horizontal strain at the point W. Again, suppose we consider W to act as a fulcrum with the weight acting at A or B—$\backslash frac\{W\}\{2\}$; the moment of this force will be $\backslash frac\{W\}\{2\}\; \backslash times\; \backslash frac\{l\}\{2\}\; =\; \backslash frac\{Wl\}\{4\}$, and the horizontal strain acting with the leverage $d$ will be $S\; \backslash times\; d$; and by equating these $Sd\; =\; \backslash frac\{Wl\}\{4\}$, or $S\; =\; \backslash frac\{Wl\}\{4d\}$, as before.

By a similar method of reasoning, the horizontal strain at any other point caused by a load in the middle can be determined.

Referring once more to the figure, we will endeavour to determine the strain at the point $W$.

Still considering the point W to act as a fulcrum, and the weight $\backslash frac\{W\}\{2\}$ to act upwards at each of the points A and B,

Iron Bridges. each half beam will act as a cantilever or as a beam fixed at one end and loaded at the other.

The weight $\backslash frac\{W\}\{2\}$ will act with a leverage $r\; =\; \backslash text\{distance\; \}\; \backslash pi\; C$ with an effect $\backslash frac\{W\}\{2\}\; \backslash times\; r$, and the horizontal strain at $w$ acts with a leverage $d$, its moment being $S\; \backslash times\; d$; therefore

which becomes, when $r\; =\; \backslash frac\{l\}\{2\}$, $S\; =\; \backslash frac\{Wl\}\{4d\}$, as before.

The requisite strength of the top and bottom flanges of a girder to support any weight at the centre is thus at once determined.

If the weight, instead of being placed on the centre of the girder, be equally distributed over its length, it is evident we still have a load equal to $\backslash frac\{W\}\{2\}$ at each end of the girder as before, but the horizontal strain at the centre $W$ is now only half what it was in the former case; it is, in fact, the same as though one-half of the entire distributed weight were accumulated at the centre, and this will be easily understood if we reflect, that although the weight at each end is still $\backslash frac\{W\}\{2\}$, each half of the beam no longer tends to fold itself around the point $W$, at the centre of the beam as before, but around a series of points equally distributed along the beam as shown in the figure.

Fig. 20.

The weights , may be considered as acting at their centre of gravity, half way between the points of bearing, and one-half only of their pressure acts directly at the centre in a vertical direction.

We may suppose the whole weight to exist in the form of two solid bars $AC,\; CB$, laid upon the beam, one-half of the weight of which is evidently supported at $A$ and $B$, pro-

Fig. 21.

ducing no transverse strain on the beam, while the other half alone acts directly at $C$.

The beam at the centre therefore suffers from an equally distributed weight only half as much as though the same weight were accumulated at the centre, therefore we have for the horizontal strain in the centre, caused by a uniform load,

And to find the strain at any other point caused by a uniform load, we have the following proportion:—As the square of the half span is to the strain at the centre, so is

the rectangle of the segments into which the given point

Fig. 22.

divides the beam to the strain at that point. Or referring to the figure, we have, expressed algebraically,

the strain at the point $O$.

For all these beams or girders a general rule appertains, which is

in which $W$ = the load, $a$ the area of the section strained, $l$ the length, and $C$ the constant as before. With respect to this rule, however, it will be necessary to observe that the constant $C$ will vary according to circumstances—that is, it will be different for beams loaded and placed in different positions.

The above rule equally applies to a solid section where $a\; =\; bd$, and $bd^2\; =\; ad$; and to show the analogy between the preceding rules for the horizontal strains and the above, we shall have for a beam loaded in the middle

and therefore, in this instance $C\; =\; \backslash frac\{4S\}\{a\}$; but if the load be uniform $C\; =\; \backslash frac\{8S\}\{a\}$; this constant will, therefore, be different for all the various applications of the load and position of the beam.

We shall now consider those simple girders which have open work sides, including the trellis and Warren girders.

If the top and bottom flanges of an equally loaded girder are united by vertical bars, the elements of a beam are wanting in such a system, and flexure takes place, as in

Fig. 23.

the figure. The diagonals , become elongated, while the diagonals , become shortened, this distortion being greatest at the ends of the beam, and nothing at the centre panel $C\_1E\_2$.

Assuming the curve to be a circle, the number of the panels to be $n$, and $T$ and $T\_1$ to be tangents at $C$ and $C\_1$; it is evident, that since the chords are equal, the angles , are all equal, and the angle $DCT$, which measures the distortion of the first rectangle, is $n$ times as great as the angle $TC\_1C\_2$, which measures the distortion of the rectangle at the centre; consequently when $n$ becomes infinite, as in a girder with plain sides, the angle

in the centre = 0, or the strain on the centre diagonals is nothing, increasing regularly towards each extremity. This result must be considered rather as an illustration than a demonstration; but to illustrate this important subject more fully, let us first consider all the verticals and diagonals perfectly inelastic, confining our attention to one panel $C\_1E\_1$. It is evident that from the compression on the top flange, the two points $C\_1,\; C\_2$ must have approached each other, and the two points $E\_1,\; E\_2$ must become further removed from each other; and this change of form of the top and bottom flange is greatest at the centre of the beam and nothing at the ends; the ends of the beam $CADB$ do not therefore remain parallel, but approach each other at the top, and recede at the bottom. Now it is of extreme importance to remark, that this distortion of the rectangle $C\_1E\_1$ from the compression and extension of its top and bottom lines does not necessarily involve any change in the length of the verticals or diagonals, or, in other words, it may take place without producing any strain in them.

If $ABCD$ is the original rectangle, $A\_1,\; B\_1,\; C\_1,\; D\_1$ will represent the distorted rectangle, the side $CD$ being compressed to $C\_1D\_1$, the side $AB$ elongated to $A\_1B\_1$, the remaining sides $A\_1C\_1,\; B\_1D\_1$ and the diagonals $A\_1D\_1,\; B\_1C\_1$ remaining precisely of the same length as in the original rectangle, but the line of junction of the diagonals is raised from $O$ to $O\_1$, and the whole depth of the beam is diminished.

The girder may therefore, from mere change in its top and bottom flanges, assume the following form, even if the verticals and diagonals were perfectly inelastic.

If, however, these rectangles are filled in with solid plates, it is evident that these plates must be stretched and compressed, and consequently distorted through every atom of their composition, before the rectangle $ABCD$ can be changed into the figure $A\_1B\_1C\_1D\_1$. Hence the remarkable stiffness and rigidity of girders with solid plates in their sides, and the small amount of deflection in such girders, as compared with those whose sides are formed of open or trellis-work; the difference resolving itself into this—that in the trellis girder deflection is only resisted by the elasticity of the top and bottom flanges, and in the girder with solid sides it is resisted, in addition to the above, by the elasticity of every particle of material of which the sides are composed.

Again, if we suppose the top and bottom flanges perfectly inelastic, and the verticals and diagonals alone capable of extension and compression, flexure takes place solely by the elongation of the diagonals , and by simultaneous compression of the diagonals , as in fig. 24.

In this case the distortion of the rectangles is greatest at the extremities of the girder, and nothing at the centre; the ends would now remain vertical, and, if plates were substituted for the diagonals, the change of figure of these plates would again involve compression, extension, and distortion of every atom of which they are composed.

We have then this important deduction, that as regards

that portion of the entire flexure of a girder which arises from extension or compression of the flanges, the trellis-work, whatever may be its strength, does nothing towards its diminution, while the plate, on the contrary, does resist this flexure; and as the distortion is greatest at the centre of the girder, the centre panel in a trellis girder might be omitted; but it is by no means useless to insert a plate on the centre panel of a solid-sided girder, though such has generally been assumed to be the case.

Again, as regards that portion of the whole flexure of a girder which arises from elongation or compression of the verticals or diagonals, it is evident that such change of form is greatest at the extremities, and whether plates or trellis-work be used, the strength of the vertical rib should at those points be greatest, while at the centre it may be reduced to zero.

Now, in practice, the deflection of the girder arises from both the above-named causes combined, viz., from the elasticity of the flanges and that of the vertical rib; and to secure the least deflection possible, it is evidently necessary to strengthen the flanges at the centre, to avoid that which arises from their change of form; also to strengthen the vertical rib at the extremities, so as to avoid the deflection caused by its distortion. We therefore perceive that trellis-work can only be substituted for plates at a sacrifice of stiffness, and that plates are efficient in insuring stiffness, even in the centre panel of the beam.

If we consider the origin of the strain in a girder, we arrive also at the same result. Let $A,\; B,\; C,\; D$ be the flanges of a girder with vertical bars between them. Half the

weight of the girder $\backslash frac\{W\}\{2\}$ is supported by the vertical prop $AC$. Then the strain on $AC\_1$ is to the vertical pressure at $A$, as the diagonal $AC\_1$ is to the vertical $AC$; the strain on $AC\_1$ varies therefore as the cosecant of the angle of inclination $C\_1AC$.

Although in a girder with plain sides, the horizontal strain vanishes at the extremities, this is not the case with the trellis sides; for the strain on $AE$ is evidently to the pressure at $A$, as $AE$ is to $EC\_1$, or varies as the tangent of the angle of inclination $C\_1AE$. If this angle be $45^\backslash circ$, then the tension on $AE$ is exactly equal to the pressure at $A$.

The strut $AC\_1$ may be replaced by the tension-rod $CE$, and with the exception that the strain will now be a tensile one, it will in every other respect be similar to the strain on $A$.

If the strut and tension-rod are both used, each may be assumed as resisting one-half of the strain that would come upon either singly.

If a thin plate is inserted, it may be considered as of no effect in resisting compression, but as replacing the tension-

rod $CE$; and with respect to its requisite strength we may arrive at a very useful practical approximation as follows:—

If $C\_1,\; E\_1$ and $C\_2,\; E\_2$ are the verticals as before, the portion of shaded plate, $C\_1\; E\_1$ takes the place of the tension-rod between $C\_1$ and $E\_1$, and being subjected to the same strain, requires the same sectional area across the lines $ab$ and $cd$. Each fibre in the direction of the shading being also of the same length as the diagonal which they replace, it is evident that the quantity of material required per panel is approximately the same whether the diagonal or the plate be employed. We have alluded to the advantage on the side of the plate with respect to flexure. It is also evident, from the figure we have given of the distortion of the plate, that this tensile strain is not the only efficient resistance it offers to the fracture of the beam, so that the strength arrived at above is certainly on the safe side.

Having determined the requisite strength of the diagonals and verticals of the extreme panel, and of the centre panel, which, in fact, need only be of sufficient strength to act as an independent beam in supporting the rolling load which comes upon it, the strain on the intermediate panels will be as their distance from the extremity. This may be illustrated as follows:—

If an equally loaded girder be cut in two at the centre, it will be perfectly restored by the insertion of a strut at C and a chain at E. The tendency of either half to descend at E is precisely counterbalanced by the similar tendency of the other half, and no diagonal is necessary to maintain equilibrium. If a weight $W$ be placed at the centre of one half, this equilibrium is destroyed, the point E now has a tendency to descend vertically, with a pressure equivalent to $\backslash frac\{W\}\{2\}$, and a strut at AC, or a tension-bar BE equivalent to this strain must now be inserted to restore stability. Hence the centre panel requires diagonals proportionate to the weight of the rolling load.

Now, if the beam, instead of being cut through at the centre, be cut through at any point CE, it is evident that a strut and a chain will no longer restore the stability of

the system, the portion of the beam EA, supposing its weight to be $W\_1$, acts with a downward pressure $\backslash frac\{W\_1\}\{2\}$ at the point E, which exceeds the downward pressure $\backslash frac\{W\}\{2\}$ of the other portion. We therefore have a true measure of the vertical force at every point of the vertical rib, which is always equal to half the difference of the weights sustained on the two sections of the beam into which the point divides it, and at once determines the requisite dimensions of the diagonals, which are consequently in exact proportion to their distance from the extremity; so that if a beam be supposed 10 feet long, and to have 10 panels, with 1 ton placed upon each panel, the vertical strains commencing from the extremity will be as follows:—

In addition to these strains, we have evidently to add the strain produced by the rolling load.

The strain on the verticals throughout is easily determined when the strain on the diagonal is known, the vertical becoming a strut when the diagonal is a tie-bar, and a tie-bar when the vertical is a strut. When a plate is used, the vertical evidently becomes a strut; and when a diagonal strut as well as a diagonal tie-bar is employed, or when the plates become very thick, as in cast-iron girders, it has evidently no duty to perform, and may be omitted. We have therefore all the elements of a perfect beam in the following arrangements (1, 2, and 3, fig. 31), in which ties are represented by dotted lines, and struts by full lines.

In No 3 of fig. 31, we have evidently two systems of struts and ties, and omitting either of them, we have the ordinary

Warren and Kennard girder as in fig. 32, which is doubtless

the simplest possible form of trellis girder that can be constructed, and of which a magnificent example exists in the Crumlin Viaduct in South Wales.

We shall, however, first describe the Newark Dyke Bridge, which, with the exception of some small and badly

designed cast-iron girders, was the first, and is still, as regards span, the largest bridge constructed on this principle. Newark Dyke Bridge carries the Great Northern line over a branch of the Trent near Newark, and was erected under the direction of Mr Joseph Cubitt.

This bridge (see fig. 34) consists of four independent girders, viz. two for each line of railway. The roadway is beneath the girders. The top flange of each girder consists of a series of cast-iron pipes butting end to end; the lower flange consists of wrought-iron links, and the flanges are connected by diagonals forming a series of equilateral triangles, and these diagonals are alternately struts and ties. The ties are formed of wrought-iron, and the struts of cast-iron: the length of each side of these triangles is 18 feet 6 inches.

inches; over the abutments the diameter is 13½ inches, and the thickness 1½ inch. They are turned and fitted into each other at the joints, where they are connected by eight bolts and nuts.

The diagonal struts and ties are connected with the top flange at the centre of every alternate casting by means of joint pins 5¼ inches diameter, passing transversely through the cast-iron pipes which are bored to receive them.

The bottom flange consists of wrought-iron links 18 feet 6 inches long, rolled in one piece of the uniform depth of 9 inches. They vary in number and thickness, increasing in total sectional area from the abutments to the centre, where there are fourteen links ¼ths inch thick. Their ends are swelled laterally to receive the joint pins by which they are connected with the diagonals.

The diagonal links are of the same form and dimensions as those of the bottom flange, as regards length and width, but are increased in thickness from the centre towards the abutments.

The top tube, which is 259 feet long, is formed of 29 lengths of cast-iron pipe; at the centre of the span, their diameter is 1 foot 6 inches, and the thickness of metal 2½

The diagonal struts are of cast-iron, the general section is that of a cross like the sign +, two sides of which are gradually increased in width from the lower to the upper end, in which is formed a jaw embracing the top tube and the diagonal links on each side of it. The joint pins pass through the two sides of this jaw (see fig. 35), through the two links and the tube, and then extend across the road to the opposite girder, and thus form part of a system of bracing between each pair of girders and over the roadway.

Iron Bridges. At the bottom of the strut there is a circular end piece (see fig. 35), 13½ inches diameter and 5 inches thick, through which is inserted the joint pin similar to that at the top, fastening together the bottom ties and diagonal struts and links, and extending in like manner across the whole width of the bridge.

The horizontal bracing by which the top and bottom of each pair of girders are connected together, and their lateral stability is insured, consists of hollow cast-iron pipes extending across the road, and through which the joint pins also pass.

These struts are cross-braced by diagonal wrought-iron bars 2½ inches wide, and connected with them by bolts.

The platform consists of a flooring of Memel fir 8 inches thick, resting on the links of the bottom tie. These links are suspended at the middle of their length by a pair of 1¼ inch rods secured to the top tube at the junction of the diagonals.

On the bearings at each end of the girders are cast-iron triangular frames, strengthened by a perpendicular rib from the apex to the base. They are braced transversely by arched ribs of cast-iron above, and by girders with bracket pieces below, to prevent any rocking motion. At the top of these frames are placed blocks of gun metal 10 inches square, upon which the ends of the top tubes rest. The whole weight of the bridge is thus supported on these frames.

The total length between the supports is 259 feet, and the depth from centre to centre of the joint pins is 16 feet. The clear span between the abutments is 240 feet 6 inches.

The total weight of iron is 244 tons 10 cwt., of which 106 tons 5 cwt. is wrought iron, and 138 tons 5 cwt. cast iron, to which must be added 50 tons for the platform, making the total weight of each bridge 294 tons 10 cwt. The cost, exclusive of the masonry of the abutments, and of the permanent rails, but inclusive of the staging for fixing and the expense of testing, was £1,11,000.

The Crumlin Viaduct.

As a further example of these girders, we have no engineering monument in this country more remarkable for lightness and novelty of construction than the viaduct, 150 feet span, which crosses the valley of Crumlin in South Wales, at an altitude of nearly 200 feet.

The lofty piers on which it is supported are equally novel, and in perfect keeping with the girders; they are composed of groups of cast-iron columns only 12 inches in diameter, cross-braced, with wrought-iron ties; and the slender and elegant appearance of this gigantic system of skilfully combined struts and ties can scarcely be imagined without seeing it. The merit of this design is due to Mr T. W. Kennard, by whom the viaduct was also erected.

The gradations are evident between the simple form of girders last described, through trellis work of greater and greater closeness until we arrive at the plate in which the trellis bars may be considered infinite in number. We shall not stop to investigate the actual deflections of such girders, which we have already seen is less as the trellis bars are more numerous, nor the effects of counter bracing on initial strain. These investigations, as well as other applications of these principles, such as in the construction of roofs, do not come within our present scope.

The great importance of the subject, however, cannot be too strongly urged upon the attention of the engineer. The evident defects which characterize most of our trellis bridges arise entirely from the want of any generally established principles as to their construction, and the too common error of believing that any number of mere tension bars, however arranged or thickly interlaced, can form an efficient vertical rib to a girder, or in any way modify

the central strain on the top and bottom flanges which depends solely on the depth of the girder, and is perfectly independent of the system which connects them.

As the best examples we can introduce of well designed trellis girders we shall select the Boyne Bridge, and some simple trellis girders of peculiarly light construction, erected by Mr Edwin Clark on the Victor Emmanuel Railway in Piedmont.

In the first timber bridges constructed in America on Boyne the trellis principle, the trellis work, and, in fact, the section of all the parts throughout, was nearly uniform. As all the trellis rods in one direction act as ties, and in the opposite direction as struts, there is thus a tendency in such beams to buckle or twist unless the trellis work is proportioned for such strain. It is evident, also, from what we have stated, that as regards trellis work the distortion will be greatest near the extremities of the beams. In fact, this has been experienced in some of our first iron trellis bridges, including the Boyne Bridge, which is the largest, and one of the earliest of these structures, though great skill has been evinced in proportioning the various parts to the strains to which they are subjected.

This magnificent bridge or viaduct was erected by Sir John McNeil on the line of the Dublin and Belfast Railway over the River Boyne, near the town of Drogheda.

It consists of three spans, viz., of a centre span of 264 feet, and two side spans, each of 138 feet 8 inches. The height above high-water spring tides to the soffit of the bridge is 90 feet.

The bridge is approached at each extremity by a series of arches 61 feet span.

The roadway is supported by two wrought-iron girders, braced together at intervals over the top. Each span is not isolated, but the whole are united into one continuous beam throughout the bridge, and every advantage is taken of the extra stiffness and strength thus obtained.

When several consecutive girders are thus united so as to form one single continuous girder, the pressure and the strains, from the weight of the bridge itself or its load, are entirely modified. Not only is the absolute deflection at the centre of the span decreased, but there are points of contrary flexure in its length, the portion over the piers presenting a concave surface at the under side, and the centre portion between the piers having a convex under surface.

The horizontal strains by this arrangement are now due only to the effect of the load on the spans between the points of contrary flexure, and the beam is virtually shortened. This shortening amounts in a perfectly continuous beam of equal spans to about two-thirds of the entire span.

In the bridge we are describing each girder has double sides formed of wrought-iron bars crossing each other at right angles, and at an angle of 45° with the horizon. The diagonal distance across the meshes, or distance from centre to centre of joints, is 7 feet 5 inches, and the total depth is 22 feet 6 inches.

The pairs of diagonal bars which sustain compression are inside, and connected together by lattice work; the tension bars being placed on the outside, and rivetted to the compression bars wherever they cross.

At the centre of the middle span the bars are 10½ inches wide by ¾ inch thick, and decrease towards the piers, where they are 4½ inches wide by ¾ inch thick.

The cross section of each beam is thus rectangular; the distance between the sides being 2 feet 3 inches, and the total width of the top and bottom flanges three feet.

The trellis sides are not attached directly to the top and bottom tables but to a vertical continuous plate, ¾ inch thick and 17½ inches broad, which is reckoned as part of these tables. The trellis work is thus only 20 feet 10 inches in depth, there being 2 feet 10 inches of side plate, which

is made to form part of the sectional area of the top and bottom chords, though really its centre of effort to resist horizontal strains can scarcely be considered to coincide with that of the actual chords.

The main girders are placed 24 feet 6 inches apart, and are braced at intervals across the top with trellis cross-beams.

The platform is 24 feet 6 inches wide, and is formed of 6 inch Memel planking resting on transverse lattice beams, 3 feet 4 inches deep at the centre, and placed 7 feet 5 inches apart.

The effective area of the upper flange of each beam is about 113.5 square inches in the centre of the centre span, and 127 inches at the lower flange. These areas diminish towards the points of contrary flexure, where they are at a minimum. They then increase again towards the piers, where the areas are at a maximum.

The following table will show the areas at various sections of each main beam:—

The total combined effective area in the centre of centre span is said to be 227 square inches, in which, as we before observed, the area of the connecting side plates is also included.

It is evident that in this bridge the centre span is not a portion of a perfectly continuous beam, the points of contrary flexure being probably about 210 feet apart.

The centre span was constructed with a camber of 4 inches in the middle, which decreased on removal of the scaffold to $1\backslash frac\{1\}\{2\}$ inch.

In testing the bridge with a load of 2 tons per foot run, the deflexion was 1.9 inch.

The weight of iron work in the two beams forming the centre span, and which carry two lines of railway, is as follows:—

Bridge over
the River
Isere.

The bridge was opened for traffic on the 5th April 1855.

The bridge constructed to carry the line of the Victor Emmanuel over the River Isere, at a point between the villages of Montmeilen and Aiguebelle, in Savoy, is on the lattice principle, in which the diagonals forming the web are placed at an angle of 45° with the horizon.

The railway has but a single line of way, and is supported by two girders, which are continuous throughout, forming a total length of 558 feet.

The bridge has four uniform openings of 130 feet $4\backslash frac\{1\}\{2\}$ inches, and is supported by stone abutments at each end, and by three piers of solid masonry in the river.

The foundations of the piers are on cast-iron cylinders, three to each pier, 6 feet 6 inches in diameter.

The thickness of the masonry is 8 feet 10$\backslash frac\{1\}\{2\}$ inches.

The level of rails is about 18 feet above the bed of the river; but the water in floods occasionally rises fully to this height, and, in the spring of 1856, it carried away a similar bridge a few miles lower down, and deposited the entire structure about 20 yards from its site in a perfect state.

The piers and abutments form an angle of 45° with the direction of the stream.

The depth of the girders is uniform throughout, being $\backslash frac\{1\}\{2\}$th of the span, or 11 feet 9 inches in extreme.

The usual width of the squares formed by the vertical gussets is 11 feet 3 inches, therefore there are twelve of these bays over each opening.

Over each pier, and at the abutments, large cast-iron frames are placed on each side of the girder to support the top flange and provide for the additional strain on the web caused by continuity; and beneath these are sets of cast-iron bed plates, between which rollers are inserted to allow of free expansion.

The diagonals of the web that have to withstand principally the force of tension are simply bars of which the width is reduced as the strains become less. For compression the struts are formed of T iron rivetted back to back, and likewise graduated in size.

The gussets or vertical struts are made up of small angle irons and plates, which latter are stiffened on their edges by light T iron.

The diagonals and centre plates of the gussets fit between, and are rivetted to the deep vertical plates and angle irons of the upper and lower flanges.

The section of the top of the girder is of the form of a T, while that of the bottom flange is an inverted T.

The horizontal parts are formed of one row of plates, with their covers and packing pieces, and the vertical parts of two plates placed far enough apart to allow of the diagonal struts and ties fitting between and fastened to the horizontal chords by two heavy angle irons.

The roadway is supported by very light transverse open girders of wrought-iron rivetted beneath the lower flanges of the main girders with which they form right angles.

They are placed 3 feet 9 inches apart, and the weight is distributed over several of them by similar open wrought-iron girders fixed in a longitudinal direction under the lines of rail.

The flooring is simply of 2$\backslash frac\{1\}\{2\}$ inches timber planking, laid parallel with the rails, and bolted to the transverse girders.

The sectional areas of all parts of the girders vary according to the calculated strains when their continuity is affected only by the weight of the permanent load is considered, but the strain under the testing load of a ton per foot run does not cause a strain of more than 3$\backslash frac\{1\}\{2\}$ tons per square inch for compression, and 4$\backslash frac\{1\}\{2\}$ tons for tension.

The greatest possible strain that can be brought upon the diagonals is much less.

The tendency of the upper part of the girder to lateral motion is constrained by a few light wrought-iron arches fixed immediately above the gussets, which being fastened to the roadway girders, the whole form strong rings or frames at such points as effectually preserve the rectangular shape of the trough. For this purpose it was necessary to use great care in fixing the main girders so that the gussets of each pier should be in a direct line.

The weight of iron work in each pair of main girders for each span was 50 tons, and of its proportion of roadway 21 tons. The total weight of iron in the whole structure is 322 tons wrought and 15 tons cast-iron.

On the same line of railway there are also two other bridges, similar in every respect, but of only a single span each.

The girders being therefore independent, the weight is slightly increased; that of the pair of girders to each bridge being 72 tons, and of the whole bridge with roadway 96 tons.

We will now give an example of a bridge with flanged girders, which also belongs to the class "Simple Girders." the Yssel. The bridge we are about to describe crosses the River Yssel, and carries the extension line of the Dutch Rhenish Railway.

The River Yssel, during the melting and breaking up of

the ice, becomes a wide and deep stream, and very dangerous. The project of a bridge to carry the railway over it was at first considered by the Dutch engineers as being so very difficult of execution, and so impossible to make secure, that, when definite plans were submitted, an enormous increase of strength in the design was required to withstand the action of the ice, which they considered would become jammed together into one mass across the whole of the river, and thus, being acted upon by the whole force of the stream, would carry away the entire structure unless extraordinary strength and solidity were used. To meet these views the girders were made on the close boiler-plate principle, and the bridge was supported by cast-iron cylinders (two to each pier) each 15 feet diameter, except the centre cylinder, which carries the roller path of a swivel bridge, and is 28 feet diameter.

The bridge has six openings, of which the two centre ones are each 50 feet in the clear, and are crossed by continuous wrought-iron girders, which revolve on the 28 feet cylinder, forming the swivel bridge with two openings. The other openings are each 164 feet in the clear.

The bridge carries a double line of railway, which it supports by three parallel girders, the centre ones being double the strength of the others. These principal girders are made perfectly continuous for the two spans on each side of the centre, and therefore their height is uniform between the centres of these spans. Since there are but two cylinders used for each pier, the centre girder is supported on "sandwich" girders, consisting of wrought-iron plates placed vertically, with wood bolted between them. By this arrangement the clear span is increased to 172 feet.

The upper flange of each main centre girder is formed of cast-iron, the extreme halves exactly corresponding, so that, if brought together, they would form an arch; but, as they are constructed, appear like an arch cut asunder, spaced apart by a parallel beam. The tensile strain in the top, induced by continuity, is entirely borne by wrought-iron plates attached to the web above the cast-iron, which extend throughout the parallel part, but vary in section proportionately to the strain; and therefore over the centre piers there is no strain whatever on the cast-iron, which, in consequence, is much reduced in size.

The lower flange of the centre girder is made up of four rows of plates, and four rows of angle iron. The web is made double, forming a box between which the cast-iron of the top flange is inserted. The side girders are simply of the ordinary I shape, but a single web plate being used, the castings of the top flange are attached outside the web, and are surmounted at the central part by a wrought-iron plate. The bottom flange is also formed of four rows of plates, of less width and thickness than those in the centre. These main girders are fixed in the centre, and the ends resting on rollers can readily expand. The roadway throughout the bridge is of 12-inch balks, placed 12 inches apart, suspended to the lower flanges of the girders. The swivel bridge turns on a central steel pivot and twenty rollers, each 2 feet 8 inches diameter, fixed to radial arms, and retained in place by a strong wrought-iron girder ring. They are of cast-iron, turned, and revolve between the planed conical surfaces of the upper and lower tram-plates. The adjustment for the bearing at the ends of the girders is by wedges moved by machinery.

The weight of wrought-iron used in the whole superstructure was 851 tons, and of cast-iron 336 tons, and the weight of wrought-iron in the lower ends of the cylinders was 64 tons, and of cast-iron 508 tons. Much of the above weight was unnecessary, and was merely added to overcome the scruples of the Dutch government engineers. Other examples of similar construction may be adduced, in which the bottom flange and web of the girders are formed of solid boiler plates, and the top flange of a combination

of wrought and cast-iron, the latter being generally in excess. Amongst those remarkable for their large span may be mentioned the bridge that carries the line of the Manchester South Junction and Altringham Railway over the River Mersey in the town of Warrington, which is of 180 feet clear span; and another, but larger structure, connecting the same line of railway with the Birkenhead, Lancashire, and Cheshire Junction Railway at Walton, near Warrington, which is parallel to the brick viaduct on the London and North-Western Railway, and also crosses the Mersey and Irwell Canal. The centre span of this bridge is 172 feet, and the others are of the respective spans of 63, 60, and 37 feet, with a bridge constructed to open of the clear span of 53 feet. The larger girders bear on 9 feet cylinders, and the smaller on cylinders of 6 feet diameter.

(4.) Trussed Girders.

It must be remembered that the only material employed in the construction of girders on their first introduction was cast-iron.

Its uncertainty and weakness when exposed to tensile strain, as in the lower flange of a girder, soon attracted the notice of engineers. Little benefit was obtained by increase of thickness; for the treacherous character of the material increased rapidly with the mass in which it is cast, and simple girders were thus limited in their dimension to very moderate spans. The difficulty of uniting cast-iron rendered impracticable the attempt to build up such girders of separate castings, and nothing remained but to attempt to strengthen the lower flange by the addition of wrought-iron tension rods. The first difficulty that presented itself was to secure a due degree of tension on the rods so employed, as the mere attachment of them to the lower member of the girder without initial strain, though it might prevent the destruction of the girder in case of the fracture of the lower flange, would evidently do but little towards preventing such fracture. It was with this object that the rods employed were attached at each extremity of the girder to its upper flange, and at the centre only brought down below the bottom flanges, and were then brought into tension. It is evident that by tightening the screws by which these rods were suspended any amount of initial strain could thus be put upon the wrought-iron ties, causing a corresponding counter strain in the girder itself; and while the rolling load was sufficient to cause a deflection equal to the counter strain thus given, no strain could come upon the lower flange of the girder itself. Provided the upper flange were sufficient to resist the thrust to which it is subject, it is evident that such girders are far less liable to accident than simple castings, and are capable of application to much larger spans.

The determination of the strength of such girders is, however, a difficult task. A serious accident, moreover, which occurred from the failure of a girder of this description at the Dee Bridge, near Chester, has entirely put an end to their employment. In these bridges, the cast-iron girder formed the whole depth of the truss, the tension rods passing beneath its lower flange; it therefore possessed considerable strength as an independent girder, without counting on any assistance from the truss. Such girders are, in fact, compound girders formed by combining the truss with the simple girder, the upper flange doing duty as a compression bar in both systems, and being thus subjected to two independent strains.

It is evident, therefore, that if the upper flange is simply proportioned to its duty as the top flange of the simple girder, it will be of insufficient strength for its additional duties as part of the truss. It has been argued, that from the perfect union of the top flange with the vertical rib, a

considerable portion of the whole girder might be taken as forming part of the truss. It is, however, evidently impossible by calculation to say how far such assistance may be relied on; and a still greater objection exists in the fact that such girders consist of two systems, the ultimate deflections of which are utterly different,—the girder, for instance, may be broken before the truss attains half its ultimate deflection, or has done half its duty. The objection to this girder is common to all girders in which two independent systems are attempted to be blended; and, as a general principle, all such arrangements should be avoided. It has always been usual, in order to obtain additional strength in the attachment of the three castings of which such girders are composed, to increase their depth where they are united as well as at each extremity of the girder, and the tie-rods, instead of being in a line with the upper flange of the girder, are attached to the upper portion of this increased depth. It is certain that in such an arrange-

ment we have no right, in calculating the strength of the truss, to count on the additional depth so obtained, and it has been contended that girders so trussed may in some cases have been actually weakened. This appears, however, to be scarcely possible, when we reflect that in all such girders as are usually constructed the tightening of the ties increases the camber of the girder. It is useless to say more on the subject of this form of girder, as since the adoption of wrought-iron for girders they have been en-

tirely superseded; they were designed when no other means existed of obtaining iron girders of great span; and the melancholy accident which occurred at Chester is the only existing instance of their failure, while the evidence given on that occasion renders it highly probable that even in that case the fracture was occasioned by the train running off the line. In the trussed girders over the River Arno, in Italy, strained tie-rods were introduced beneath the lower flange from end to end of the girder, and the experiments made on this girder were highly satisfactory.

One example of this system exists in the bridge carrying the road from Banbury to Lutterworth over the London and Birmingham Railway (fig. 37). Span 64 feet.

There are six ribs which have a double curved form, or rather that of a parabolic spindle; the lower curve being formed by a wrought-iron tie-bar, and the upper one by a cast-iron arch; the space between the two is filled by cast-iron ornamental panels, which are made to act as struts, and for the purpose of keeping the bottom tie-bar in the proper line of curvature and strain. The tie is in two parts, with eyes at each end, and pins passing through them, which take the ultimate strain. A very ingenious arrangement is applied at the centre of the rib for the purpose of adjusting the degree of tension upon the tie-beam. It consists of a vertical tubular strut, attached to the top arch by a screw, which is also used to force it down upon the tie. On the top of the arch there is a kind of plate-band cast with it, which forms the level for the cast-iron road plates. The structure is exceedingly tasteful in appearance; and the nice arrangement in the parts makes it something more of a mechanical contrivance than is generally understood by an iron bridge.

As an example of these bridges, we shall describe one of the earliest and largest, viz., the trussed girder bridge of 63 feet span, erected by Mr Bidder, for carrying the Blackwall Railway across the Minories in London. This structure, as well as all the bridges on the line, was not originally designed to carry locomotives, the line being

worked by a rope. It is fortunate that in this, as in most of our early railway works, sufficient excess of strength was given to allow of the greatly increased weight of our present railway traffic as compared with what was then anticipated. The bridge consists of 6 girders, viz., two outside girders and 2 girders beneath each line of rails. Each

girder is formed of three separate castings with upper and lower flanges. The upper flange is 8 inches wide, and the bottom 2 feet wide. The general depth is 3 feet, but the depth is increased at each extremity of each casting to 4 feet 6 inches. The joints by which they are bolted end to end is thus 4 feet 6 inches deep. The joint is further strengthened by wrought-iron clips beneath the bottom flange.

The trussing rods, by which the tensile strain on the lower flange is relieved, and by which an initial camber was given to the girders, consist of wrought-iron bar 5 inches wide and 1 inch thick, placed in pairs on each side of the girders. These links are attached to the top flanges of the girders at their extremities, and descend diagonally to the bottom flanges at each joint, where they are connected with each other, and with the girder by a pin passing through the latter. They are tightened up by means of keys at each end of the girder.

The bridge has subsequently been widened by moving one of the outside girders so as to admit of the introduction of another line of rails, the girders being strengthened by the addition of top pieces which are firmly bolted to the upper flange, making the total depth of altered girders 5 feet.

After the accident which occurred at Chester, similar additions were made to many then existing girders, an additional top flange being inserted in a direct line between the attachment of the tie-rods.

The above girders resemble in principle another ordinary form of trussed girder used for travelling cranes and other

purposes where light girders of considerable span are required, timber being frequently used for compression, and wrought-iron tie-bars for tension. The depth in such cases is generally obtained by the insertion of one or more light cast-iron standards or struts between the tie-bar and the timber. The principles on which the strength of such girders is calculated are perfectly identical with those given at page 592, inasmuch as the centre of the timber strut may be taken as a fulcrum, and the strain on the rods depend upon the distance from this point. The stiffness of such girders is of course greatly increased by the addition of diagonals between the standards, and in this

form, when made of durable materials, they become excellent girders for all the purposes where moderate spans only are required.

We have, however, two examples of bridges of enormous dimensions constructed on the same type, viz., the Chepstow and Saltash Bridges, both of which have been Bridge erected by Mr I. K. Brunel, and both of them equally remarkable for their gigantic proportions, and the engineering difficulties which had to be overcome, not only in the superstructure, but in their foundations.

The Chepstow Bridge (fig. 39) carries the South Wales Railway over the River Wye at a height of 46 feet above high water. The remarkable rise of the tide which characterizes the Bristol Channel is well known, and at Chepstow is no less than 41 feet; at high springs the elevation of the bridge above low water is thus nearly double. In this bridge two kinds of girder are employed,—one half of the bridge consists of ordinary wrought-iron girders of 100 feet span,

resting on cast-iron columns; in the other half, the roadway is carried over the river by the trussed girders, which have a span of no less than 305 feet. This portion of the bridge is in fact a rigid suspension bridge, the tension of the chains being resisted not as an ordinary bridge by attachment to the ground at each end, but by a horizontal cylindrical wrought-iron column, or strut, 9 feet in diameter and $\backslash frac\{1\}\{2\}$ inch thick, which rests on the towers at each end

of the bridge. Instead of the ordinary catenary, the chain consists of three straight links only. The rigid form of the chain is preserved, and the flexure of the horizontal column is prevented by their mutual attachment by diagonal and vertical bracing, the girders which carry the roadway are suspended from the chain at two points only, viz., at each end of the centre link. These girders are thus divided into three nearly equal spans, and are supported at each end by cast-iron.

The tower at the extremity of the bridge rests upon the precipitous rock which bounds the river, but at the other end upon a pier consisting of six cast-iron columns, which pass through 50 feet of mud down to the rock beneath. The mode of sinking these cylinders was novel. They were placed in their position on the site of the pier, which is dry at low water, and the mud was excavated till they began to sink with their own weight, when fresh lengths were added on the top as the previous lengths sank down. They were thus ultimately bedded on the rock, and filled up with concrete. The cylinders are carried up to

a height of 190 feet, and are connected at the top by the cast-iron framing and tower which carries the tubes.

The weights of the various parts are as follows:—

The tubes or struts are of uniform section throughout, and are formed of sixteen equal plates $\backslash frac\{3\}\{8\}$ inch thick, and two side plates $\backslash frac\{1\}\{2\}$ inch thick. The plates are all 10 feet long, lapped together at sides, and butt-jointed at the ends with double butt plates, and rivetted together with two rows of rivets.

(5.) Bowstring Girders.

The difficulty of obtaining abutments capable of resisting the thrust of large arches, more especially when the use of

iron in their construction allowed of great diminution of their rise or versed sine, led naturally to the addition of a chain or tie to resist their thrust. Again, the outline of ordinary girders with parallel top and bottom flanges, when of any magnitude, is by no means agreeable. Depth being the only visible element of strength, the eye does not fail to perceive that such an outline implies equal horizontal strains throughout, where we feel that the strains at the ends of a beam merge entirely into a vertical direction, and that depth is there useless. In large girders the depth at the centre was therefore alone increased, and the top took an arched form, the curve being generally a parabola; the analogy of such a girder with the bowstring arch is at once apparent. It is, however, of great importance to understand thoroughly the difference between these two systems.

The strain at the centre of each of them is, as we have before seen, perfectly identical. The weight supported at the ends is also in both cases equal to half the weight of the girder; but the manner in which these similar vertical forces are converted into similar horizontal forces between their origin at the points of bearing and their mutual equilibrium at the centre of each girder, is in each case entirely different, and the horizontal strain in the top and bottom flanges varies accordingly.

In an uniformly loaded ordinary girder, with parallel top and bottom flanges as we have already shown, the horizontal strain varies at any point of the flanges as the rectangle of the segments into which the point divides the span. We must, therefore, in order to insure equal strains throughout, either diminish the section of the flanges in that proportion as we recede from the centre, or we must diminish the depth of the girder in a similar proportion, preserving uniformity of section in the flanges. In either case the strain per square inch remains constant throughout the whole length of the flanges; but, as their actual section at corresponding points is entirely different, it is evident that the actual strains to which they are subjected are also entirely different. Now, as these strains arise entirely from the action of the vertical rib which connects the flanges, it is evident that the duties of the vertical rib are entirely different in the two cases.

In the parallel-sided girder (fig. 40), in which the upper and lower members taper off to 0 at the extremities, the plate A' requires sufficient stiffness and strength, both vertically and diagonally, to support half the weight of the girder, and will be the thickest side plate in the girder; but in the parabolic-shaped girder (fig. 41), where the flanges maintain their full section to the end, the plate A'' requires no such additional strength, and may be reduced to nothing. In fact, if we now omit the vertical rib altogether, fig. 41 is a bowstring girder. These girders, therefore, have little analogy except as regards the strain at their centres, or, which alone is identical on both; the difference between the two girders is evident.

In fig. 40 the horizontal strain in the flanges is at a maximum at the centre, and decreases to 0 at each end, and the section varies in a similar proportion.

In fig. 41 the horizontal strain on the flanges is constant throughout their length, and the section is also uniform throughout.

In fig. 40 the vertical strains from the point of support are thrown gradually into the flanges by the action of the ver-

Iron Bridges. tical rib. They are nothing at the centre, and at a maximum at the ends where they are entirely resisted by the vertical rib.

In fig. 41 the vertical strain is thrown directly into the flanges without the intervention of any vertical rib.

Now, a bowstring girder is such an arch, with horizontal ties to resist its thrust in lieu of abutments, and as regards supporting its own weight, if the arch be one of equilibrium, and the ties are suspended from the arch to preserve their horizontal position, no diagonal bracing or vertical rib of any kind is necessary. We may, therefore, at once apply all the principles applicable to the ordinary arch in investigating the strains of such a girder.

It is evident, however, in the same manner as with the arch, that such a structure in equilibrium would not support any unequal load. The best means of giving rigidity to this skeleton arch is a problem on which much ingenuity has been expended. There are three methods by which the rigidity has been obtained,—1st, By giving sufficient rigidity to the arch itself, or the ties themselves, or both, to ensure the requisite stiffness as in the High-level Bridge at

Newcastle; 2d, By cross-bracing between the arch and the ties as in the Monkland Canal Bridge; and, 3d, By an independent system of trussing or framing, combined with the arch and ties as in the proposed girders for the Mayence Bridge, and in most American timber bridges. By any of these additions, or rather by a combination of them all, it is probable that this system of girder is capable of very great extension. An example of each of these bridges is here given.

The earliest railway bridge on the bowstring principle is that over the Regent's Canal, near Chalk Farm, on the London and Birmingham Railway (see fig. 42).

It is composed of three main ribs of cast-iron open panel-work, whose outline is parallel, but which includes an arc extending to its extremities of length and depth, and intersecting the vertical bars which form the panels. The span is 50 feet and the height of the ribs 10 feet. The section of each rib is in the form of a hollow rectangle 2 feet 11 inches wide, and the space between its sides is filled with diagonal bracing-frames 5 feet 10 inches apart.

The railway is carried by cast-iron girders of the fish-

Fig. 42.—Regent Canal Bridge.

bellied shape, 28 feet between bearings and 1 foot 10

inches deep in the middle; they are suspended from the bracing-frames in the main ribs by wrought-iron suspension rods $2\backslash frac\{1\}\{2\}$ inches diameter; there being sockets in the bracing-frames to receive their upper ends, and in the ends of the cross-girders to receive their lower ends.

The centre main rib performs double duty; and its bracing frames have double sockets, and carry two suspension rods. In addition to the ribs themselves in resisting the strain of the load there are longitudinal tie-bars under each rib, there being four under each outside rib in a horizontal row, and eight under the centre rib in two horizontal rows. These tie-rods are secured to the bearing ends of the main ribs, and are in three lengths, each united by sockets, gibs, and keys. Upon the cross girders are oak sleepers for the rails; and the entire space between the rails is filled in with cast-iron plates perforated in the form of trellis-work. The outsides of the outer ribs are ornamented with cast-iron mouldings and fret-work. This bridge is of very bold design, and certainly a novelty as regards construction.

The finest example of the kind of structure we are alluding to is undoubtedly the High-level bridge at New-

Fig. 42.—(Transverse Section.)

castle-on-Tyne. This bridge (fig. 43), which crosses the River Tyne, unites the towns of Gateshead and Newcastle. The Tyne runs through a deep valley or ravine, on each

side of which these towns are built. The old bridge crosses the river at the bottom of the valley, and the want of a bridge at the high-level was long severely felt by the in-

habitants for the accommodation of their local traffic. So far back as the commencement of the present century, various plans were projected to meet this requirement; but

the great cost of the undertaking rendered it at that time impossible to carry any of them into effect, and they were reluctantly abandoned.

The traffic between the two towns, however, rapidly and constantly increased, especially when Gateshead became the terminus of the Southern Railways, which were thus entirely isolated from the lines north of the Tyne; some other means of communication between the termini at Gateshead and Newcastle became now, in fact, indispensable, the cost of conveyance of passengers and merchandise by coaches and omnibuses across the old bridge reaching the enormous amount of £1000 per week.

In a work so indispensable for both interests the local authorities gladly co-operated with the railways, and it was at length determined to construct the present bridge, and to carry both a public road and the railway across it.

There are two platforms, the upper platform carries three lines of railway, while the lower forms the common public road.

The breadth of the river at this spot, at high water, is 515 feet, but the whole distance between the Gateshead station and the central terminus on the Newcastle side is about 4000 feet. The approaches of the railway (as will be seen by the engraving) are curves in contrary directions, but those of the public road below are straight.

The bridge is in six spans, each of 125 feet, and the superstructure is supported on stone piers and abutments, at a height to the soffit, above high water, of 83 feet.

The foundations consist of piles, the spaces between which are filled up with concrete. Many of the piles are 40 feet long, and all are driven through the hard sand and gravel, forming the bed of the river, until they reach the solid rock.

Many difficulties occurred in driving the piles which considerably retarded the progress of the work, and, among others, the peculiar effect of ebb and flow during this operation is worthy of note. At flood-tide the sand became so hard as almost totally to resist the utmost efforts of driving, while at ebb the sand was quite loose, and allowed of doing so with facility. It was therefore found necessary to abandon the driving on many occasions during high water.

The difference between high and low water is 11 feet 6 inches.

Another difficulty arose from the quicksands beneath the foundations. Although the piles were driven to the rock bottom, the water forced its way up, baffling the attempts to fill in between them; this, however, was remedied by using a concrete made of broken stone and Roman cement, which was continually thrown in until the bottom was found to be secure.

The piles were driven by Nasmyth's Steam Pile-driver, Nasmyth's this being one of the first cases in which this valuable Pile-engine was used, a foundation was thus obtained, which certainly would not have been possible by the ordinary means.

The ram weighed a ton and a-half and had a fall of 2 feet 9 inches. It was worked incessantly, night and day, driving at the rate of sixty or seventy strokes per minute. In several instances the pile heads burst into flame, and burnt fiercely under this rapid action of the ram.

In setting out the spans previous to the driving of the Guage permanent piles, guage piles were driven with the ordinary pile-engine as deep as they would go. When the steam pile-driver was introduced, an experiment was tried with it upon one of them, which was driven to a farther depth of 15 feet.

One of the foundation piles was tested with a load of 150 tons which was allowed to remain several days, and upon its removal no settlement whatever had taken place. The piles are 4 feet from centre to centre, and the utmost that can come upon one of them is 70 tons, supposing none of the weight to be carried by the intervening space of planking and concrete.

The foundations of the abutments are upon a bed of strong clay, under-lying the sand and gravel at a depth of 18 feet; no piles are used; the abutments are of stone similar to the piers. The roadway is carried for a considerable distance upon each side of the bridge upon masonry arches of very durable construction. Designs for the proposed

Iron Bridges. entrances to the bridge, and in keeping with the architectural features of the structure were designed by the engineer, but have not yet, for financial reasons, been carried into effect.

The cofferdams for the piers (none being used for the abutments) were formed in double tiers, filled in with puddle. The piles were not drawn, but were cut off level with the bed of the river by a circular saw, the lower part remaining to protect the foundations as well as to avoid disturbing the bottom by their extraction.

Stone work. The foundation course lies about 2 feet below low water. The stone is of a hard and durable quality, and came from the neighbourhood, viz., a portion from Heddon-on-the-Wall, and the remainder from Corbridge.

Dimensions of piers, &c. The dimensions of the piers are as follows:—

The most novel features, however, in this structure are the bowstring arches over the centre portion of 900 feet. This space is crossed by six similar spans. The four centre spans cross the river, the remaining two are on the slopes on either side.

Each bay is crossed by four main arched ribs with horizontal tie-bars to resist the thrust. The upper roadway rests upon the arches, the lower is suspended from them by wrought-iron suspension rods.

Each arch is cast in five segments, strongly bolted together, and when entire is 125 feet in span, with a rise of 17 feet 6 inches from the centre of the tension bars, and of 18 feet 1½ inches from the upper surface of the bed plates. The depth of the arch at the crown is 3 feet 6 inches, and at the haunches 3 feet 9 inches. The section is that of a double-flanged girder, the flanges being 12 inches wide, and 2 inches thick on the outer arches, and 3 inches thick in the internal arches, which have a greater weight to support. The vertical ribs are of the same thickness as the flanges.

The sectional area of the external rib at the crown is 133 inches, and of the internal ribs 189 inches; the combined area of section at the crown in the four arches being 644, and at the haunches 706 square inches.

The ties consist of flat wrought-iron bars, 7 inches by 1 inch of best scrap iron, with eyes of 3½ inches diameter, bored out of the solid, and pins turned and fitted closely.

Each external rib is tied by four of these bars, and each internal rib by eight. The sectional area of each external tie is 28 inches, and of each internal tie 56 inches, giving a total area of 168 square inches.

These bars were all tested to 9 tons on the square inch.

The four ribs are disposed in pairs (see cross section); the two internal ribs being 20 feet 4 inches apart in the clear for carriage road, and the space between the internal and external ribs is 6 feet 2 inches, forming a footpath on each side.

The arched ribs are braced between the footpath space by cast-iron vertical and longitudinal bracing-frames, with 2-inch tie-bolts passing through them, which extend as near as possible to the haunches, allowing only a sufficient headway for the passengers.

The spandrels between the arches and the beams which carry the railway, are filled with cast-iron vertical square pillars, also braced across with diagonal bracing frames, similar to those between the ribs. On the tops of these spandrel pillars are trough-shaped longitudinal girders, which extend along the bridge as continuous beams.

The ribs have square bosses cast on them, corresponding with the spandrel pillars, and forming horizontal tables for

them to stand on; these were all truly bored and faced.

The spandril pillars are continued downwards, concealing the suspension bolts, and add great stiffness to the superstructure.

On the longitudinal top-girders are the cross-girders, also of a trough form, cast in one length, extending across the whole four ribs. The cross-girders have pockets or shoes cast on them, to receive the ends of the longitudinal timber-bearers upon which the road-planking is spiked.

All butting joints were planed on their upper surfaces to receive the bearing ends of the ribs, which were also planed.

On the abutments all the bearings are allowed to expand and contract, but on the next bearings they are fixed down firmly with four 1½-inch bolts and nuts; on the next pier again they are free, and so on alternately fixed and free.

There are no rollers, but the bearings have sliding surfaces provided for the purpose.

The motion caused by expansion and contraction, ascertained by observation, was, during a variation of temperature of 32 degrees, .92 inch in the whole six spans—153 inch per span.

Each arch was temporarily erected at the contractor's works, and tested before removal, and all the other detailed parts received a separate test previously to this final trial.

The planking is 3 inches thick, jointed and tongued with hoop iron, and laid in two courses crossing each other at right angles. The upper course is caulked and pitched with as much care as would be bestowed on a ship's bottom.

The lower roadway is formed of cross planking in a similar manner, and is paved with wooden blocks 4½ inches cube of a new construction. The blocks are cut at the upper surface, in form of a shoulder (fig. 44), so that when they are all in close contact they show so many grooves, which extend to half their depth (fig. 45), the channels being an inch in width.

The surface on which the blocks were laid was covered with pitch; the blocks were dipped in hot pitch and laid, and the grooves filled with broken stone and gravel; a layer of pitch, and finally of sand was spread over all. Six years' wear has made but little impression upon this paving.

The quantity of masonry was,—

The cost of the bridge was,—

The first heavy portion of the superstructure was cast in February 1847, and the bridge was opened by the Queen in September 1849.

As an example of one of the simplest forms of bowstring Monkland girders which have been constructed, we shall describe the Canal large wrought-iron girders designed by Mr Edwin Clark Bridge for replacing a timber bridge over the Monkland Canal on

Fig. 44.

Fig. 45.

the Caledonian Railway. The laminated timber arches of the old bridge were always much distorted, and as in many similar bridges which were constructed on our early railways, were beginning rapidly to decay, from the play of the timbers and the infiltration of water into the joints and between the planks of which the arches were composed. Such arches are undesirable for exposed structures in a damp climate, though peculiarly adapted for roofs where the arches are sheltered from the weather, as in many stations on the Caledonian line, and more especially on the magnificent roof of the Great Northern Railway station at King's Cross. The railway passed over these arches, which, from their dangerous state, were shored up, and it was of the utmost importance to replace them by some other structure without interfering with the traffic over more than one line at a time. Wrought-iron bowstring girders were therefore used on account of their lightness and the facilities they gave for adding but little weight to the old bridge during their construction upon it. The arch or top member of these girders is partly wrought, and partly cast-iron, and this construction has been found advantageous and economical. A similar combination has been adopted by Mr Clark in the bridge at Arnheim over the Yssel, on the Walton Viaduct, the large girder-bridge at Warrington, &c.

The lower flange or tie consists of wrought-iron plates as in the bottom of an ordinary flanged wrought-iron girder. The stiffness of the tie is secured by a deep vertical rib which forms a portion of its effective section, and which renders it sufficiently rigid not only for supporting the roadway, but also for assisting materially in preserving the arch from vertical change of form. This is further secured by a similar vertical rib which forms a portion of the effective section of the arch. The vertical and diagonal bracing which connects the arch and the tie are attached conveniently to these vertical ribs. The arch consists of three ribs, the centre being nearly double the strength of the external ribs. The total length of the girders is 148 feet, and the depth is about $\backslash frac\{1\}\{3\}$th of the length, or 15 feet.

The duty of the diagonal bracing is simply to preserve the form of the girder under a rolling load, and in order to insure a sufficient connection between the arch and the tie

at each extremity, the panels at the extremities are filled in with close plates. The transverse timbers which carry the roadway, and which are simply bolted up beneath the lower flange, are 8 inches deep, and are placed close side by side, running across the whole bridge. Such a roadway is renewed with extreme facility, and occupies the least possible depth. The arches are braced together over the roadway.

The whole weight of these girders for the double line is only 128 tons, and their substitution for the old bridge did not occasion the least interruption to the traffic. The sectional area of the top flanges is 120 square inches, viz., 60 square inches of wrought-iron, and 115 square inches of cast-iron. The sectional area of the lower flange or tie is 150 square inches.

As examples of the third method of securing vertical rigidity in such girders we may refer to most of the large timber viaducts erected in America, where the arch is combined in numberless ways with horizontal trussed or trellis girders with great ingenuity and simplicity of detail. Wrought-iron arches of a somewhat similar description have been designed by Mr Clark for the bridge at Mayence on the Rhine.

In this case the arches have a versed line of about one-tenth of their length, and are connected over several spans as well as rendered rigid by a continuous wrought-iron trellis girder of half the depth of the arch which is in section a wrought-iron rectangular box.

This roadway girder is extremely light, its duty being only to distribute the weight of the rolling load. It offers great facilities for the erection of the arches, and secures them from lateral motion by partially filling in the spandrels over the piers. It also breaks the extremely undulating outline which independent arches would present. It is moreover evident that great benefit is derived from the continuity thus given to the various spans. It is believed that, in cases where the roadway can be placed on the top of this continuous girder to allow of cross-bracing beneath it, this system is peculiarly applicable to spans of great magnitude.

There is a peculiar bow-string bridge erected on the Gloucester and Birmingham Railway, by Captain Moorsom (fig. 46), which is worthy of notice.

Fig. 46.—Cheltenham Bridge.

The roadway is carried by two main ribs of 45 feet span, each of which is in form of an arch with an arched chord; the top member is 9 inches deep by 3 inches in thickness, and the lower member or chord is one foot 6 inches deep, and is of the double-flanged form. The whole rise of the top member of the rib from the springing is 7 feet 3 inches, and that of the lower member or chord is 1 foot 6 inches; the space between them at the centre being 4 feet 3 inches.

Each rib is in five castings; and the top and bottom members are tied together by vertical bolts, 2 feet 5$\backslash frac\{1\}\{2\}$ inches apart.

The roadway plates are carried on a series of cast-iron beams on the skew, one end resting on the walls, the other in shoes, provided on the lower member of the ribs. Three of the cross beams bear at both ends on the walls; the beams themselves are curved with a camber of 1 foot 6

inches, and are tied together with bolts. The whole platform is carried by roadway plates of cast-iron, on which the road materials are placed.

(6.) Tubular Girders.

In the ordinary plate girder we seem to have arrived at the limits of simplicity of construction. A top and bottom flange with a vertical rib, all composed of plates, and united by angle irons, form a beam, the calculation of the strength of which is within the reach of every one. As soon, however, as we exceed very moderate limits, difficulties arise entirely beyond the reach of theoretical research, which are in fact quite independent of the action of such a beam as a girder, but arise from the distortion of form to which such a system, supported only on two distant and limited

bearing surfaces, must as a whole be liable, more especially as expansion and contraction prevent any rigid attachment even on these contracted supports.

It is in meeting such difficulties in large structures that the skill and judgment of the engineer is really taxed when he exceeds the limits of actual experience. In the design of the Britannia Bridge, it was the mere arrangement of materials to resist the transverse strain which composed the difficulty of the problem. It was rather the practical construction of any such structure at all; the difficulty of obtaining the materials required, or of adopting such as were obtainable to such new purposes, and of devising a beam not merely of sufficient strength for its ultimate use as a bridge, but of sufficient independent rigidity for retaining its form not only when in place, but during its erection on its temporary scaffold, its flotation on unstable pontoons, and the ultimate raising of it into its place suspended isolated from four simple chairs.

The tubular girder alone seems really adapted to such varied strains; and it is difficult to conceive even now, with all our subsequent experience, any other means than those adopted of solving the great problem which thus inaugurated a new epoch in the history of bridge construction, and led directly not only to our present theoretical knowledge of the principles of beams, but to all those numberless elegant and ingenious practical combinations of wrought-iron in bridge construction, which it has been our province to describe.

These remarks on a subject now becoming old, are necessary, because great misapprehension exists as to the real objects in view in the construction of the first tubular bridges, and which, though probably the only constructions possible under such circumstances, and unsurpassed in point of simplicity of construction or economy by any other kind of beam, are yet not necessarily of universal application. The chief examples existing at present of tubular bridges of any magnitude are the following:—

1. 1. The Conway Bridge.
2. 2. The Britannia Bridge.
3. 3. The Brotherton Bridge over the River Aire.
4. 4. The bridge over the Nile.
5. 5. The Victoria Bridge over the St Lawrence.

In all these cases, a tubular bridge simply means a pair of girders, with their top and bottom flanges of sufficient horizontal extension to meet and combine, forming a rec-

tangular tube or trough, through the interior of which, or upon the top of which, the roadway may be laid.

It may here be stated, that the result of experiments made with that object had directly proved that which might at first sight appear problematical, viz., that with such extensive horizontal development of the top and bottom flanges, the whole of their sectional area acts effectively in resisting extension or compression throughout the entire width; and, in fact, in cases of absolute fracture, where such beams have been broken in the experiments, the tearing asunder of the bottom plates actually commenced at about the middle of the tube, and not at the outside edges. The whole of the principles applicable to simple girders are thus directly applicable to these bridges, and full details of these principles and of their application, with a minute description of the construction and erection of the Britannia and Conway Bridges, together with an account of the extensive series of experiments made in connection with these works, and of the practical deductions thus arrived at, will be found in Mr Edwin Clark's work on this subject, to which the reader has before been frequently referred. We shall here merely give a few details, and draw attention to some of the circumstances which must be taken into account in considering these works as simple beams.

The Britannia Bridge, which carries the Chester and Holyhead Railway over the Menai Straits (see figs. 47 and 48), consists of two independent continuous wrought-iron tubular beams 1511 feet in length, and weighing 4680 tons each, independent of the cast-iron frames inserted at their bearings on the towers. They are 15 feet wide, and vary in depth from 23 feet at the ends to 30 feet at the centre. They rest on two abutments and three towers of masonry, at a height of 100 feet above high water. The roadway is laid along the bottom, viz., one line of rails in each tube. The centre or Britannia tower, which is altogether 230 feet high, is built on a rock in the middle of the Straits. The bridge has thus four spans, viz., two spans of 460 feet over the water, and two spans of 230 feet over the land. On each side the weight of a single span of 470 feet is 1587 tons, and of a span of 242 feet 630 tons. These tubes repose solidly on the centre tower, but repose on roller beds on the land towers and abutments. Now, these gigantic dimensions are by no means the only remarkable features in this work. The opponents of the Holyhead Road had imposed conditions on the Chester and Holyhead Railway, which were thought insurmountable with respect to this bridge. The navigation

Fig. 47. Britannia Bridge—(Part Elevation)

was not to be interrupted—no scaffolding could thus be used—and the clear height of 100 feet was to be retained throughout; arches being objected to unless the springing and not

the centre was at this elevation. The tides set through this portion of the Strait with a velocity of 9 miles per hour, and the quiet water at each turn of the tide lasts but for a

Iron Bridges. very short space of time. The tubes were designed to meet all these requirements; they were so constructed at a

considerable distance from their permanent site on the shores of the Straits, they were floated upon pontoons upon these rapid tides to the base of the towers, and they were then drawn up by hydraulic presses to their required height. They were here united through the towers by the insertion of shorter lengths, and ultimately brought into the conditions of continuous beams as regards strain, by the means employed for their junction. It is evident such structures would be designed specially for such varied circumstances, for example,—

As soon as they were completed on temporary platforms, these platforms were removed, and they became isolated beams; the ends were accordingly strengthened with cast and wrought-iron framing for this special object, and had they always remained there the sides might have been throughout considerably lighter than they are; they now weigh nearly 40 per cent. of the whole weight. But in the next operation, that of floating, the tubes were liable to be supported at any point of their length, besides being subjected to chances of considerable distortion, and to disasters which on more than one occasion did actually threaten their entire destruction. The stiffening frames and gussets, which in an ordinary girder would have only been necessary at the ends, became therefore necessary throughout the whole length, and even the top and bottom were considerably modified, as it is evident that while overhanging the pon-

toons on each end to the extent of 70 feet, that the top, instead of being in compression was thrown into extension; the weight of the tubes was consequently much increased by these arrangements. Again, they had to be raised by being suspended freely from four chains. Provision for this suspension from such limited attachment had also to be made of a totally opposite character from that made for their vertical support when on their bed; and, ultimately, when raised to their place, they remained no longer independent beams, but were converted into continuous beams,—parts before in tension being now thrown into compression, and vice versa; while the ends which were before subject to no horizontal strain were now exposed to greater strain than even the centre of the span. And, last of all, during the act of raising one of these enormous masses, the press from which it was suspended burst, and one end of the beam fell through a space of no less than 9 inches on to a loose uneven heap of planks beneath it, bulging in the bottom plates, breaking all the castings, distorting seriously the sides and stiffening frames; while the broken press itself, which descended from a height of about 100 feet above, broke through the top plates and completed the crippling of the whole section of support. It may surely be doubted whether anything but a tube could have stood such unexampled violence; and in proportioning the parts of a structure destined for such usage, the mere consideration of the strain to which an ordinary beam it would be subjected, formed but a part of the problem; no direct comparison can therefore be made between the weight of this bridge and an ordinary beam. If this were the case with the large spans, it is still more so with the small spans of 230 feet, which as simple beams would weigh only 230 tons each, whereas their actual weight is 650 tons. But it must be borne in mind that as regards the bridge itself these small spans were not required at all, and that they were merely designed and used as counterpoises for the large tubes, for the important purpose of converting them into continuous beams by their overhanging weight. By examining their detail, it will be found they are designed solely for this special purpose, their use as beams being made entirely subsidiary.

Some misapprehension exists on the object and importance of the cells of which the top and bottom of these tubes is composed. These cells are rectangular, there being eight of them in the top and six of them in the bottom, and they run throughout the bridge. With respect to their importance, it must be observed that the whole section of the top of the Britannia tube at the centre is 64825 square inches, and of the bottom 58543 square inches, and that the tube is 15 feet wide; the thickness of a single plate to ensure this section would therefore have been 27 inches for the top, and 23 inches for the bottom; and had such a plate been procurable, nothing better could have been desired, and the cells would be unnecessary. Such a thing, however, is evidently impossible, and the engineer in this, as in numberless other details, had to adopt what he could obtain; now the arrangement of the plates in cells is almost the only conceivable arrangement possible for obtaining the required section, allowing access, at the same time, to every part for construction and future maintenance. This alone led to their use in the bottom of the tube, where their form was totally unimportant. With respect to the top, however, it was of great importance, since thick plates could not be had, to ascertain the best form of cell for resistance to compression that could be devised with thin plates. A series of valuable experiments by Mr Eaton Hodgkinson led to the use of the rectangular cells as actually used, not because such form presented any peculiar advantage over any other form, as some have imagined, but because these experiments demonstrated that cells of that magnitude and thickness were independent of form, and are crushed only by the actual crushing of the iron itself; under these cir-

cumstances, the square cells were used as the best practical method of obtaining the sectional area required.

Similar misapprehension also exists as to the considerations which led to the rectangular form of the tubes themselves. Now, the result of direct experiments made with round, oval, and rectangular tubes—there being precisely the same section and weight in all three, and, consequently, different depths—was, that the circular tube was the weakest, and the oval tube the strongest, the rectangular form being intermediate. The oval tube was, indeed, first studied with a view to its use. Its form, however, was not favourable, neither for its actual construction, nor for its connection with the suspension chains, which were originally intended to be used in the erection; and practical considerations, in this case, also compelled the use of the rectangular tube. It must also be remarked, that the result of experiments made on round, oval, and rectangular wrought-iron tubes, when reduced to the same depth and compared, was in favour of the rectangular form, although within ordinary limits the form was not proved to be a matter of very great importance. It may be added, that this bridge has now been in use six years, that the deflection has been carefully tested, from time to time, with the utmost precision, and that not the slightest perceptible increase has taken place during that period. The care with which the painting has been attended, and the protection afforded by the roof, have also entirely preserved it from the slightest damage by oxidation; and it is difficult to conceive that even the lapse of centuries can in any way affect such a structure, or to doubt that it will remain one of the most durable, as it certainly is one of the most remarkable monuments of the enterprise of the present century.

The Conway Bridge is in most respects similar to the Britannia Bridge, the local peculiarities of site being nearly similar. It consists of only two tubes of 400 feet span, placed side by side, and weighing each 1180 tons. The same provisions, as regards its strength, were made for floating it to its place and raising it; but, as the span is single, and it is not a continuous beam, the general arrangement of the plates is entirely different on the two bridges.

In the Brotherton Bridge, on the York and North Midland, the span of which is nearly identical with the small spans of the Britannia Bridge, viz., 225 feet, it was possible to compose the top of a single plate, and no cells whatever have been used, either on the top or bottom.

The depth of these tubes is 20 feet, and their width between the side plates is 11 feet.

Each tube rests on two sets of rollers at one extremity, the other being bolted down to the pier. The rollers are placed at the extreme outsides of the tube, immediately under the sides, extending inwards only 2 feet 3 inches.

The roller-plates are bedded on creosoted timber. The bearings are stiffened by cast-iron standards, three on each side, firmly bolted to the side plates. Each standard is about 40 inches in sectional area at the bottom, and tapers to about 30 inches section at the top.

The rails are laid on longitudinal timbers, which are supported at intervals of 5 feet by angle iron brackets, riveted to the cross beams or keelsons.

The weight of one tube is as follows:—

The rigidity of the tube exemplifies remarkably the advantage of solid or close sides in diminishing deflection. A circumstance also occurred in the construction of this bridge which illustrates one of the great advantages peculiar to tubes, viz., their independent strength.

The form given to these tubes is not rectangular, but

the top was narrower than the bottom, so that their section

was slightly pyramidal; their width at the bottom was 11 feet 10 inches, and at the top only 11 feet. Now, after the opening of the bridge, the width at the level of the carriage windows was objected to by the government inspectors, although they had previously sanctioned the width on the first tube erected. It became necessary, therefore, to widen the tubes, and this was done in a very interesting manner, viz., by literally opening the top of the tube down a centre line throughout its length, and inserting in the opening a longitudinal plate, 10 inches broad, from end to end. In this manner the sides were evidently moved farther apart to a less extent at the level of the carriages. By these means, and the removal of a portion of the projecting rib of the T irons, the whole was sufficiently widened. No other kind of beam could evidently have retained its form during so extraordinary an operation.

The principal feature in the Egyptian Railway Bridges is, that the road is carried upon the top of the tubes, and not in their interior.

There are two tubular viaducts upon the Egyptian Railway. The larger one crosses the Damietta branch of the Nile near Benha, and the smaller one crosses the Karrineen Canal at Berket-el-Saba (Lake of the Lion). These viaducts unite two old roads, formerly connected by a ferry, and each is contiguous to a vice-regal palace.

In the larger viaduct there are ten spans or openings, the

two centre ones comprising one of the largest swing-bridges that has been attempted.

The total length of the swing-beam is 157 feet; it is balanced at the middle of its length on a large central pier. When open to the navigation a clear water-way is left on either side of the central pier of 60 feet. Each half of the beam sustains its own weight as a cantilever, 66 feet long.

The eight remaining spans are 80 feet in the clear, arranged four on each side of the centre portion; and the total length of the viaduct between the abutments is 865 feet.

The piers consist of wrought-iron cylinders, 7 feet in diameter below the level of low Nile, and 5 feet diameter above that level. They were sunk by a pneumatic process to a depth of 33 feet below the bed of the river, through soil of a peculiarly shifting character, and are filled in with concrete.

There are six of these cylinders in the central pier which supports the swing-bridge; and the adjacent piers on either side of the centre have each four cylinders; each of the remaining piers has two cylinders only. The tops of the cylinders are covered by cast-iron circular plates which rest entirely upon the concrete, special care being taken to prevent any contact with the cylinders. On these circular plates rest the upper cast-iron plates which connect the piers, and form a seating for the bearing-plates of the beams.

The beams or tubes are 6 feet 6 inches deep, and 6 feet 6 inches wide at the bottom, tapering to 6 feet wide at top, and they rest at their ends on rollers working between planed surfaces to admit of the motion caused by expansion and contraction.

The tubes carry a single line of way on their tops, the rails being laid on longitudinal sleepers, and there is also a roadway 4 feet wide on either side, supported by wrought-iron brackets bolted to the sides of the tube.

These roadways are of corrugated iron, resting on the brackets, and stiffened by strips of bar-iron placed transversely on the top.

The six cylinders for the central pier are also provided with cast-iron circular plates, as before described, and surmounted by a framework of cast-iron, uniting the tops of the cylinders, and serving as the lower tramway for the rolling machinery.

The revolving machinery consists of a turn-table containing eighteen accurately turned conical rollers, their angle being determined to the greatest nicety, and corresponding with the angular surfaces of the tram-plates between which they revolve.

The diameter of this turn-table is 19 feet from centre to centre of the rollers.

The whole of the rollers, together with the wrought-iron circular frame to which they are attached, form an independent system, usually termed the "live-ring," held in its position by the central pivot. The frame of the "live-ring" is connected with the rollers by radial spindles with gun-metal gudgeons at the periphery and centre. And, to prevent any difference in angular speed between the rollers and centre portion, a very excellent arrangement is adopted,

which consists in a diagonal strap passing over the central wheel, and extending to the outer periphery. This strap is keyed up to any adjustment in which it firmly keeps the radial spindles.

The swing-tube is firmly attached to the upper tram-plate by a system of cast-iron bracket-work and strong bolts and nuts; forming, in fact, as is most essential at this point, an exceedingly rigid attachment. The centre pivot is of forged iron, 9 inches diameter, and turned accurately to fit its bearings. To insure a firm fixing for this pivot, it is made to pass through the entire depth of the lower tram-plate into a socket provided for the purpose, in which position it is firmly keyed. The bridge is turned with facility by a capstan worked by two men, with gearing communicating with the large rack surrounding the lower tram-plate.

To prevent accident to the swing-bridge when open, "Fenders" are placed up and down stream, similar in construction to the piers of the bridge. At the bearing ends of the swing-bridge arrangements are made for locking it in its position. These consist of fixed inclined planes attached to the under surface of the bearing ends of the tube and corresponding wedges which slide on the piers, which are made to recede and advance by means of a screw turned by gear-work.

In the Birket-el-Saba Viaduct the swing portion forms spans on each side of 43 feet, and the fixed portion consists of two spans of 70 feet each. In other respects the viaducts are precisely similar.

Both were commenced in May 1853, and completed for traffic in October 1855.

It may be here remarked that the duties of a swing-bridge in resisting strain are greatest when the bridge is open and only sustaining its own weight. This will be readily understood by considering the entire swing-bridge in the condition of a beam supported in the middle, and loaded at the ends, the action being, in fact, similar to that illustrated by fig. 17. The centre point in the figure may represent the centre pier of the swing-bridge; therefore it is evident that the depth must be proportioned to the entire length of the swing-beam and not to the mere span on each side of the centres; and by the simplest reasoning it will be found that the load of a railway train, if amounting even to one ton per foot, will not produce so much strain on the bridge when it is resting at its ends on the outer piers, as the beam will produce itself when open and overhanging; for in the former position, the load, though probably double the weight per foot run of the beam itself, is acting on a span supported on each end, equal to less than half the length of the beam, whereas in the latter case the weight of the entire beam is acting at a span which gives it more than twice that leverage.

There are numerous other iron railway bridges of which we could give examples, but those already given comprehend the whole of the principles of construction adopted. The engraving (Plate II.) of some of the most important bridges will give perhaps a more comprehensive idea of their magnitudes. (R. S.)