to this from above was along the right bank to within about 1,000 feet of the bridge, from which point they approached the opening at an angle, because of having to move to the left about 500 feet. The spans to the left could not be used because of the middle-ground bar, We have here an example of a bridge with the widest draw-openings on the Mississippi River, being one of the most difficult to pass, and furnishing but one practical opening for anything. In high stages the shape of the shores and bed are such that the flow through the draw . cannot be much different in direction, but at this time the dike and bars above were submerged, and the flow was more direct through the spans to the left of the draw, making it then the main channel of the river. At the time of the high water in 1876, the first very high water since the building of the bridge, the increased velocity due to the contraction by the bridge scoured out the bottom, particularly under the short spans next to the left bank. This deepening was discovered by those in charge of the bridge. They at once commenced to place riprap about the piers and along the axis of the bridge. This was being done in October, and our examinations showed that nearly the whole bottom between these piers was covered with riprap, and about the piers it appeared above water. The depth between the piers on the axis of the bridge was from 22 to 31 feet, and below the bridge the depth increased from 50 to 57 feet within 150 feet of the piers. The bottom of the piles on which the piers rested were only 27 to 30 feet below the surface of the water, so that the bottom had scoured out below them on three sides of the piers. Notwithstanding their efforts (without any considerable rise in the river), the second pier from the left abutment disappeared on the night of November 6, carrying with it two spans of the bridge. This shows that riprap protection against scour is not efficient when thrown from the bridge. The effect, indeed, is to prevent Scouring where the stone falls; but in this case the danger was from the hole made below the axis of the bridge, which the stone did not reach, and which hole finally became so enlarged as to ingulf the whole mass situated above it. To prevent scour in such instances, boats are required from which the stone can be thrown directly into the place where the material is being removed. The effect of the dike built from the left bank has not been what was expected, and the depth of water below, near the left bank, has increased. This is probably owing to the material in the left bank being much finer than on the right, and it does not seem an unreasonable apprehension that if this dike should be carried away at time of high water the navigable channel might take down the left bank, leaving the draw-openings impassable at low water. The location of the bridge just below an important landing, as in the case of Burlington, is also objectionable.

ALTERATIONS TO LESSEN OBSTRUCTIONS TO NAVIGATION.

The great narrowing of the water-way by embankments at this bridge forms an exceptional case, and it was a part of the original design. When this plan was submitted to me in May, 1873, I made no objection to this feature of it, but said:

All experience, on this point, on the Mississippi shows such contraction in width is compensated for by deepening until the area of section and mean velocity are the same as before.

This could not be true if the bridge was tected by riprap from scour at high stages. time should be allowed for succeeding high S. Ex. 69-9

built at low water and proAfter any such contraction, waters to effect the enlarge.

ment before any obstruction is placed in it; not only should time be allowed for this, but the river for a mile or more above should be narrowed and the banks rectified, so as to gradually bring the conditions of the natural river to those at the proposed bridge. This was the course pursued at Saint Louis, to which reference was made by me; there the contraction and rectification of the banks extended through considerable distance, and was gradually effected in the course of a number of years. Such sudden narrowings, as at Louisiana, with riprap protections to prevent scour, produced very abnormal conditions of engorgement, and the want of proper controlling works above left us without the means of giving permanence to the directions of the currents or the position of the channel. The only thing to do is to supply what has been omitted. In the first place, it is certain that if we do not deepen the section at the bridge to an extent too great for the safety of the present structure, we must widen the water-way. In the modifications which I propose, the Noix Creek span is given to that stream exclusively, and will not be taken into the following consideration of increasing the section of the bridge.

The present section is at low water 1,750 feet wide, excluding the piers; depth, 103 feet; area of section, 18,200 square feet; volume, 47,000 cubic feet per second; and a mean velocity of 2.48 feet per second.

It is proposed to increase this width by removing a portion of the embankment on the Illinois side, and by building two spans of 250 feet each in the clear; this will increase the low water width to 2,250 feet. We shall then have a new section with:

To remove and prevent the reformation of the sand-bar in the middle of the river above the bridge, and to give proper direction to the cur rent, it will be necessary to build new banks and cross-banks running out from the present shore-lines to them. The proposed new shore on the right bank will be on the prolongation in nearly a straight line of the extension of the right rest-pier up to the landing at Louisiana. This new bank will entirely shut off Noix Creek. If it is necessary to use the mouth of this creek for any purpose it must be done by coming into it from below the bridge.

The line of the proposed new left bank will extend from the proposed new abutment in a line parallel with the new right bank up river about 4,200 feet. The distance between these banks is about 2,400 feet. The cross-banks should be from 300 to 600 feet apart, as experience shows necessary. The work should be begun by building these cross-banks. These should be commenced at the shore, and extended outward by

laying down mattresses and weighting them. The bottom layer should be put in over the whole distance at once to prevent scour at the ends, and they should be raised at such rates as the observed effects call for. The portion of the present wing-dam above the bridge, between the proposed new banks, should be removed. It is unnecessary to enter into details of this operation here. The frequent failure of these crossbanks or jetties in this country has been due to their being placed too far apart. There may be no necessity of building the new left bank more than to put a T at the end of the cross-banks; and their heights there need not exceed the height of the present bottom land. The new longitudinal bank on the right side should, however, extend 4 feet above the highest water, so as to produce no cross-currents into or out of Noix Creek.

Similar regulation of the river to that proposed here is used at Kinlenburg Bridge, in the Netherlands. The difficulty of making a suitable draw-bridge at this point will probably lead to the adoption of a high bridge in its place.

Sheer-booms.-The board of engineers recommended as an immediate alleviation of the difficulties experienced in passing rafts, that the bulkhead in extension of the right rest-pier be continued up river at least 500 feet, and this line extended 820 feet further by a fixed boom, its upper end 300 feet from the shore at ordinary low water.

HIGH BRIDGE.

A high bridge at this place would have to be placed 57 feet abovehigh water. It would have an easy approach from the right bank and a high approach on the left bank. It should have at least three high spans, with clear openings of not less than 300 feet each.

SAINT LOUIS ELEVATED RAILWAY AND HIGHWAY BRIDGE.

(Opened for travel in July, 1874.)

Authority. This bridge was built under provisions of sections 11 and 12 of the act approved July 26, 1866, and act approved July 20, 1868.

DESCRIPTION OF THE RIVER AND VALLEY AT THE LOCALITY.

The Mississippi has been subjected to great artificial contractions at Saint Louis, beginning in 1838 when dikes were commenced to prevent the enlargement of the channel east of Bloody Island and improve the one on the other side. These works were added to from time to time, and in 1850 a low water stone dam was built across the channel east of Bloody Island, which, however, was afterward carried away. In 1852 the present high water dam connecting Bloody Island with the Illinois shore was built and made a highway. Prior thereto this channel had become so filled up that water did not run through till the river was 13 feet above low water. (See report of Col. W. E. Merrill, major engi neers.) The site of Saint Louis is mainly above overflow. The shore in places is of rock. Considerable portions of the river front have been graded, sloped, and paved with stone. Bloody Island, now East Saint Louis, has been partly raised above overflow, and the river front also sloped and paved.

The low lands in Illinois known as the "American Bottom," about 6 to 8 miles wide, are protected from overflow. The whole volume of the liver is therefore confined to a narrow channel about 1,200 feet wide at

low water and 1,900 feet at high water. The slope of the water surface, at both high and low water, is on an average about .65 foot per mile. The maximum velocity and volume have not been determined. The ordinary high floods are about 35 feet; the greatest known 41 feet above low water.

The city directrix is 33 feet above low water. The scouring in time of floods and ice-gorges is very great. The bed-rock lies about 90 feet below the surface of low water on the Illinois side, and probably underlies the whole "American Bottom" at about that level. Surveys were not made by me at this place, all the important information to be obtained in this way having been determined by other officers of engineers. The wide spans and piers parallel with the current give ample space on the water, but the headway is insufficient to allow the large steamboats to pass it at high water. (See discussion in Chapter III.)

DESCRIPTION OF THE BRIDGE.

The information available for a description of this bridge is contained in a report of the engineer-in-chief in 1868, and the modifications of it given in his report in 1870, together with the discussions in regard to it, printed in the transactions of the American Society of Civil Engineers. It is probable that we shall, in due course of time, have a full account of the bridge in detail, and of all the operations connected with it; but it has such an important relation to the navigation of the Mississippi that it is thought best to attempt to give a general idea of it here. The abutments are placed on the margin of the water at ordinary low stage, with two piers in the channel-way. The three spaces thus left are spanned by circular segment arches. The middle one has a chord of 515 feet and versed sine of 472 feet. The lowest part of the arch at the springing line is 71⁄2 feet above the city directrix, or at the level of high water of 1844. The crown of the arch is therefore 55 feet above the city directrix. The chord of the side spans is 497 feet, and the versed sine 43 feet 10 inches. The springing lines at the abutments are 18 inches lower than at the piers, so as to lower the crown about 9 inches, leaving the height at that point 50 feet above the directrix. Mr. Theodore Cooper says (Trans. Am. So. Civ. Engrs., vol. iii, p. 239):

Each span is composed of four arched ribs spaced respectively 16, 12, and 16 feet apart from center to center (Figs. 3 and 4). Each of these ribs is composed of two lines of tubes (called the upper and lower member according to their position), spaced 12 feet apart between centers of the tubes. The several tubes are about 12 feet long. The upper and lower members of each rib are connected together by a system of triangular bracing. The first or skew-back tubes of each member are rigidly fastened (screwed into) the large wrought-iron skew-backs, which in turn rest upon the bedplates set into the masonry, the whole being firmly anchored to the piers or abutments by large steel anchor bolts. The several tubes forming one line or member are rigidly connected at their ends by grooved couplings (Fig. 5), carefully fitted to grooves cut on the ends of the tubes. The couplings are made in halves with flanges for bolting them together. Through the center of each coupling and the ends of the tubes passes a taper pin-hole for the steel pin, upon which the braces and cross-stays are attached. All joints were made as nearly perfect as the best mechanical appliances would permit. The section of the tube is a circle. Each tube has an envelope of steel one quarter of an inch thick. There are six steel staves in each tube. Their thickness is as follows, beginning at the springing lines: Staves in first tube, 2 inches; second tube, 17 inches; third, 1; fourth and fifth, 17 inches; thence to the center 13 inches.

16

The two lines of tubes, forming a single vertical rib, are united to gether and braced by a single triangular system of vertical bracing,