memorable one in Europe. According to the incomplete accounts received, the destruction in Northern Germany had been especially marked. The sluggish flow of the rivers, the many channels which some of them occupy in parts of their courses, and the exposed condition of the adjacent low-lying lands were considered as occasions, among others, of the frequent floods of the region. But a very significant cause was to be found in the prevailing northward course of the rivers. In consequence of this the snow upon the mountains about the headwaters often melts rapidly while the ice near the mouths of the rivers has not begun to yield, and may even continue to form. Hence the spring freshets of the southern highlands discharge both water and floating ice upon the frozen rivers of the northern lowland plains. There the still firm ice upon the rivers obstructs the progress of the floating ice, thereby forming dams which cause the waters to accumulate and overflow or break through the river embankments and inundate the surrounding territory.

The speaker contrasted the fortunate southward flow of the New England rivers with the disastrous northward course of the German streams. He further remarked upon the effect of forests around the upper tributaries of southward-flowing rivers in retaining a considerable amount of snow until late in the spring, thus furnishing a protection from floods and contributing to the value of such streams for water power and for navigation.

THE DEVELOPMENT OF Bridge building.

The Chairman then introduced Prof. George F. Swain, of the Institute, who gave a brief sketch of the development of bridge building. The paper was accompanied by a large number of lantern views of ancient and modern bridges, complete or in process of construction, illustrating the different points mentioned.

Prof. SWAIN said: Notwithstanding that all of us are interested in the subject of bridges, popular ideas in regard to them are very vague. I am often asked, for instance, “which is the stronger, an arch or a suspension bridge?" or whether "a cantilever is a strong kind of bridge." Questions like these show that the general public do not appreciate or understand the methods followed by engineers in designing such structures. In designing a bridge, the loads to be carried are first determined, The bridge is then built so that

it shall require a load five or six times as great as the actual load to break it down, or, as we say, with a factor of safety of five or six. It will at once be asked how we can possibly have bridge disasters if bridges are built to carry five or six times the loads actually put upon them. It might seem to show engineering science to be lamentably insufficient, but the fact is that it is extremely rare for a bridge to break down from faults of construction, and when one does, it is always from extreme faults of design or material which could easily have been avoided.

I shall endeavor tonight to present to you a very brief and necessarily incomplete account of the development of bridge building, which may render clear some points often misunderstood; and at the close I shall show you some illustrations of faults of design which will show the necessity of allowing this margin or factor of safety.

The first bridges of which we have any account were of stone, and these were built long before Christ, by the Egyptians, Assyrians, and Romans. They were for the most part arches, usually semicircular, or nearly so; and by the Romans the construction of stone bridges was brought to a high degree of perfection. At that time stone arches were used on a large scale principally in connection with works of water supply, although highway bridges were also built; and there was in Rome a bridge across the Tiber, with a span of eighty-four feet, and others of smaller dimensions. Rome was supplied with water through nine aqueducts, the first two of which were built under ground, so that, in case of invasion, the city should not be deprived of water. The third, however, was built partly above ground, and so strong that the two succeeding ones were built directly above it, so that they had in places three tiers of arches, one above the other, each carrying a conduit. Arches with more than one tier were also built where but one conduit was to be carried. For instance, the Pont du Gard, which carried across the river Gardon the aqueduct supplying the city of Nismes, had three tiers of arches, and this construction was probably adopted on account of the ease with which the conduit could be maintained and inspected. This bridge, built in the time of the Emperor Augustus, was 885 feet long on top, and 157 feet above the stream. The longest arch in the lower tier had a span of 80 feet 5 inches, while others had spans of 63 and 51 feet. The arches of the upper tier had all the same span, 15 feet 9 inches.

From the time of the Romans until the Middle Ages little was done in bridge building, although the Goths, in Spain and Italy, built some stone arches, among them the remarkable aqueduct of Spoleto, with piers of over 100 meters high. In 1454 a segmental stone arch was built in France, with a span of 179 feet; and in 1599 a very bold arch was constructed in Nuremberg, with a span of 96 feet, and a rise of only 13 feet.

With the introduction of railways bridge building received an enormous impetus, and stone bridges, as well as other bridges, were constructed in great numbers in various parts of Europe. They were sometimes built with tiers of arches, one above another, but the spans nowhere exceeded about 200 feet. For works of water supply, as well, a number of large works of this kind were executed. One of the largest of these bridges is the Roquefavour Aqueduct, carrying the conduit which supplies the city of Marseilles with water. This structure is 1287 feet long, 262 feet high, and has three tiers of arches. Between 700 and 800 workmen were employed upon it for seven years, and the total cost was $750,000.

Although many bridges of similar character, for aqueducts and railways, may be found in France and Germany, there are in America comparatively few stone arches, and on railways they are rarely found, except for very short spans. The Thomas Viaduct on the Baltimore and Potomac Railroad, crossing the Patapsco River, on a curve, with eight elliptical arches of 58 feet span, was one of the earliest stone bridges in this country. Another was the Starucca Viaduct, on the Erie Railroad, 110 feet high, with 18 arches of 50 feet span; and a third crosses the Schuylkill River just above Philadelphia.

In connection with works of water supply, however, stone arches are not unfrequently met with in this country. The largest of these, and the largest existing stone arch, is the Cabin John Bridge, carrying the Washington Aqueduct, built in 1866, with a single span of 220 feet, and a rise of 574 feet. As other and familiar examples may be mentioned the Harlem Bridge of the old Croton Aqueduct, 100 feet above the Harlem River, with seven spans of 50 feet and eight of 80 feet; and the Charles River Bridge of the Sudbury Aqueduct, which supplies the city of Boston, the principal arch of which has a span of 127 feet with 42 feet rise.

Stone arches, while expensive in first cost, have the great advantage of permanence. Many of the old Roman arches are standing today, and in good condition. A stone arch requires but little expense of maintenance, and is not liable to be outgrown as the weight of rolling stock increases, for the reason that the principle upon which the stone arch is built is essentially different from that governing the design of any other kind of structure. Other bridges are designed so as to be strong enough to carry given loads. If these loads are exceeded, the margin of safety is reduced. The question of stability has scarcely to be considered. A stone arch, on the contrary, is designed primarily so as to be stable, so that it will not tumble down, as an arch built of loose blocks may do. The question of strength must, of course, be consid ered; but, fortunately, if a stone arch is stable, it is almost always amply strong even for much larger loads than are generally to be put upon it. A stone arch, therefore, if once correctly designed so as to be stable, is not generally liable to be rendered dangerous even if the weight of rolling stock is doubled.

Wooden bridges first came into very extensive use on the railroads of this country; and, in fact, they have been used here more. than anywhere else. One of the earliest types was the Town Lattice, patented in 1820, and built entirely of plank, joined together with oak treenails. Another early type was the Howe Truss, patented in 1840. This was the first truss in which iron and wood were combined, and it and the Town Lattice are still the standard wooden bridges. Among the early wooden bridges may be mentioned the Portage Viaduct, on the Erie Kailroad, a wooden trestle bridge, about 230 feet high, which was entirely destroyed by fire, May 6, 1875. Another was the Cascade Bridge on the Erie Railroad, an arch with a span of 275 feet and a rise of 45 feet, probably the largest wooden arch ever built.

Wooden bridges may easily be made amply strong up to spans of 150 or 200 feet, and they have some advantages; but their first cost, for large spans, is not sufficiently less than that of an iron bridge, to make up for their shorter life, greater cost of maintenance, and the added danger from fire.

The earliest iron bridges were of cast iron. The first of these was commenced in 1755, in Lyons, and was intended to be three arches, each with a span of 25 meters. It was, however, not completed, the

construction being changed to wooden girders; and it was left for England, the birthplace of the iron industry, to produce the first iron bridge. This was a cast-iron arch, built between 1773 and 1779, across the River Severn. at Colebrookdale, with one span of 100 feet. It consisted of three concentric arches, connected by radial pieces, the inner arch being cast in but two parts, which were joined at the crown. The difficulty of making such large castings induced an Englishman named Payne to experiment with small hollow castings, in the shape of arch stones; and between 1793 and 1796 a cast-iron arch, with oue span of 236 feet, and a rise of 34 feet, was built by Burdon over the Wear, at Sunderland, on this principle. It was composed of ribs, each built of cast-iron voussoirs put together just as the voussoirs of a stone arch are put together, and held in place laterally by three wrought-iron rods on each side of the rib or arch ring. Several bridges were built on this same system, one of which, over the Thames at Stains, collapsed; and in 1801 the noted engineer, Telford. projected a cast-iron arch to cross the Thames, at London, to be built on this system, with the extraordinary span of 600 feet.

The next step was to construct cast-iron arches as a series of larger castings, united by flanges and bolts, an example of which is the Southwark Bridge over the Thames at London, built by Rennie, between 1814 and 1819, with three spans, the longest of 240 feet, with a rise of 24 feet. Cast-iron arches were also built in France and Germany on each of the two systems described. Sometimes the castings were in the shape of tubes, and were united by flanges and bolts.* As another example of the reliance placed upon cast iron as a material for bridges, in those early days, it may be mentioned that the first plan made by Stephenson, in 1844, for the Britannia Bridge, was for a cast-iron arch, to be composed of pieces bolted together, the span to be 475 feet.

Cast iron, besides being used in arches, was also used in the shape of beams of short span; and it was furthermore common, for many years, to employ it in combination with wrought iron in the same bridge. Some of the early plate girders, for instance, were built with two webs, the top flange being of cast iron, to which the webs were bolted, while the lower flange was, of course, made of

It may be mentioned here that an arch of this kind exists in Washington, being composed of water pipes, through which the water flows to the city.