quantities of coal, ores, grain, &c., unequaled in the world. The transportation by means of barges will undoubtedly increase, and no restriction to its development to the fullest extent should be allowed by bridges or anything else. Rafting. The raft interest, which is of the utmost importance to the country in supplying cheap lumber, is now carried on by steamboats in the same manuer as the coal-fleet system, and with much greater economy than formerly, when manual force at long oars was relied upon to guide these cumbrous floating masses. This business, too, requires wide spans, much greater than those provided by existing draw-bridges, and high enough to allow the steamboats to pass under and guide the raft through. Large Upper Mississippi rafts are about 270 feet wide and 450 feet long, and all such have now to be made so as to separate in two parts to pass the draw-bridges on the Upper Mississippi. (See, further, report of Board of Engineers on sheer-booms; Annual Report Chief of Engineers for 1877, p. 817; also appendix to this report, contained in House Ex. Doc. No. 41, Fortyfourth Congress, second session.) General remarks.-When the bridges across the Mississippi were authorized in 1866, the barge system of transportation had already made itself important there, and although the rafting by steam was not then practiced, the inapplicability of draw-bridges was seriously felt. Railroads running in the direction of the river, however, were not yet developed, and passenger and mail transportation along it was of such great importance, that some persons much interested in the navigation thought that draw bridges with 150 feet clear openings were superior to high bridges which gave head way of only 50 feet at high water. The building of railroads has diminished the proportionate amount of passenger travel on the river, and increased the amount of bulky and heavy products and manufactures, and has already changed these relations. The building of the bridge at Saint Louis, giving a headway of only 50 feet at ordinary extreme high water, has compelled a change in the boats running above, so that they are now prepared to pass under any bridge without trouble or delay which gives the same clear height and has spans of 300 to 400 feet width in the clear for rafts and barges, so that one permanent high bridge without draw fixes a limit of height for all places above, and when a boat is built to pass under one she can without more trouble pass all such. On the contrary, each draw-bridge causes its own obstruction, and the total obstruction increases with the number, which is now so great and liable to cause such deteution as to prevent all regularity in running a line of steamboats. Experience on the Upper Mississippi has shown that the proper site for a bridge is where the channel is straight and next to the bank of both high and low water. The channel spans, whether low draws or high fixed ones, would then be next the bank, which at all times would furnish the best guide for the pilot; and then the further reason that the side next the shore can be protected by piling or other means that will allow the vessels to rub against it with safety; and this location also permits of auxiliary works being built out from the opposite shore to make the channel permanent. Such locations cannot readily be made where the rise of the river from low to high water much exceeds 20 feet, because the necessary slope of the banks will require an open span between the margin of the low and high water shores. It is therefore not generally applicable to the Ohio, or Missouri, or to the Mississippi, below the mouth of the Illinois River. On none of the Western rivers has nature prepared such a channel of navigation as upon the Mississippi above the junction of the Missouri. Its gentle slope and current, clear water, freedom from snags, limited range of height of floods, all render its navigation easy; and its condition is expressed to the eye of the traveler in its aspect of beauty. Its very gentleness has led to its being bridged with very trifling structures, which a turbulent stream like the Missouri would not suffer to stand. The latter has exacted much better bridge constructions on the average than those existing on the Upper Mississippi. The navigation of the Upper Mississippi has a much more extensive interest than that represented by Saint Louis and the Upper Mississippi, considered by themselves. This navigation is but a part of a system which includes the whole Mississippi basin, the Ohio, the Missouri, the Illinois, &c. The vessels which navigate the Ohio should be allowed to carry their loads to Saint Paul as well as Saint Louis, and any structure which compels a change of character of vessel on the upper and lower sides of it, or a breaking of bulk, is simply an obstruction to navigation of the character of the rapids at Louisville, at Keokuk, and at Rock Island, which the government is spending large sums to remove. In view of the probability that there will be before long a necessity for rebuilding some of the present draw-bridges at least, and substituting high ones, I have thought it proper to present the gauge data I have been enabled to get affecting the question of height of bridge when draw-spans are omitted, and discuss the question of headway under bridges. This has never been done, or even practicable until recently, for want of observations. This subject will form the remainder of this chapter. Knowledge of varying stages important.-The surface of the water of the Mississippi is generally rising or falling, with occasional periods when it is stationary, and a knowledge of the annual average duration of these various stages is one of the important factors in considering the effect upon navigation of bridges of different heights at differeut places. Gauge data used.-With the view to gain this knowledge we have obtained the most important continuous gauge-readings along the river, from the earliest records, which were made during the period between 1860 and 1876. This period does not include any very remarkable floods, such as those of 1828, 1851, and 1858; but in some places it does the remarkable low waters which occurred in 1863 and 1864. In the following presentation and discussion, records of gauge-readings have been obtained and used in comparison At Saint Louis, from January 1, 1861, to December 31, 1876. At Quincy, from January 1, 1869, to December 31, 1876. At Rock Island, from January 1, 1861, to December 31, 1876. At Clinton, from October 17, 1866 and 1867, while river was open. At Dubuque, from May 1, 1869, to December 31, 1876. At Prairie du Chien during 1867, '68, '69, while river was open. At Saint Paul during 1867, '68, '69, while river was open. The observations at Saint Louis are kept at the expense of that city, part of which were furnished us direct, part were derived from the report of the board of engineers on the Saint Louis bridge, and part through the assistance of Maj. C. J. Allen, United States Engineers, in General Simpson's office. Those at Quincy were obtained from Chicago, Burlington and Quincy Railroad Company, through Mr. W. Beckwith, superintendent of bridges. Those at Rock Island are from observations by the old bridge company up to 1869, and since then from observations made under direction of the engineer officers of the Army. Those at Dubuque were furnished by the bridge company, Hon. William B. Allison, president, through Mr. C. H. Booth, secretary. All the foregoing named were obtained without expense to the survey. The others were made for the purposes of this investigation and at its expense. The gauge-readings made at Keokuk might have been obtained, but its situation at the foot of the rapids destroys their general value, and the canal makes a bridge at this place a special one in character, to be determined by the necessities of the canal navigation. There were other gauge observations attainable besides those named, but were not useful in the general consideration of this subject. Gauge ob ervations at Saint Louis and Rock Island compared for period of 16 years -The longer period of continued observations at Saint Louis and at Rock Island makes the averages deduced from them more reliable than at any other points, although the situation of both these places is peculiar. At Saint Louis the Mississippi is immediately affected by its long tributary, the Missouri River, and at Rock Island by the rapids. These peculiarities are shown in Diagram H, in which the average readings of the gauges are taken for the period, 16 years, between January 1, 1861, and December 31, 1876. The gauge-readings at Saint Louis are divided by 2, so as to bring the two curves nearer together on the diagram. This does not give an equality of rise and fall, and therefore they are placed with the midway point in common, so as to divide the excess about equally at the times of high water in April and May, and that of the low waters in Novem ber and December. We might have made an exact equality of rise and fall by reduction, but this is near enough for this comparison. The peculiar differences show for themselves on the diagram. For instance, at Rock Island there is a rise in December, January, and February, not so marked at Saint Louis, which is probably due to the ice at Rock Island. With the spring rains there is a considerable uniformity of rise at the two places, but it reaches a greater relative height at Rock Island, and much sooner subsides. As a matter of further interest, we have prepared Diagram I for this 16-year period for Saint Louis, and Diagram J for Rock Island, giving the mean curve and the year when the highest and lowest water occurred on any day of the year. These show the extreme ranges that have been observed at the same dates in these 16 years, and the included space forms what we have called a zone of river oscillations, which includes all our known observations. The three exceptional high waters of 1844, '51, '58, recorded at Saint Louis, are noted on the diagram, but not included in the zone. We could not give each year's curve without too great confusion or resorting to colors which photolithography would not reproduce. These diagrams furnish interesting data for comparison and study, and information as to probability of high and low water at any time of the year, useful in planning engineering operations. The daily observations are preserved in permanent records and can easily be procured when desired for any special purpose. Comparison of means of gauge-readings at Saint Louis, Quincy, Rock Island, and Dubuque.-In order to see the effect of the natural conditions at Saint Louis and Rock Island, we have compared the simultaneous superintendent of bridges. Those at Rock Island are from observations by the old bridge company up to 1869, and since then from ob servations made under direction of the engineer officers of the Army. Those at Dubuque were furnished by the bridge company, Hon. William B. Allison, president, through Mr. C. H. Booth, secretary. All the foregoing named were obtained without expense to the survey. The others were made for the purposes of this investigation and at its expense. The gauge-readings made at Keokuk might have been obtained, but its situation at the foot of the rapids destroys their general value, and the canal makes a bridge at this place a special one in character, to be determined by the necessities of the canal navigation. There were other gauge observations attainable besides those named, but were not useful in the general consideration of this subject. Gauge ob ervations at Saint Louis and Rock Island compared for period of 16 years -The longer period of continued observations at Saint Louis and at Rock Island makes the averages deduced from them more reliable than at any other points, although the situation of both these places is peculiar. At Saint Louis the Mississippi is immediately affected by its long tributary, the Missouri River, and at Rock Island by the rapids. These peculiarities are shown in Diagram H, in which the average readings of the gauges are taken for the period, 16 years, between January 1, 1861, and December 31, 1876. The gauge-readings at Saint Louis are divided by 2, so as to bring the two curves nearer together on the diagram. This does not give an equality of rise and fall, and therefore they are placed with the midway point in common, so as to divide the excess about equally at the times of high water in April and May, and that of the low waters in Novem ber and December. We might have made an exact equality of rise and fall by reduction, but this is near enough for this comparison. The peculiar differences show for themselves on the diagram. For instance, at Rock Island there is a rise in December, January, and February, not so marked at Saint Louis, which is probably due to the ice at Rock Island. With the spring rains there is a considerable uniformity of rise at the two places, but it reaches a greater relative height at Rock Island, and much sooner subsides. As a matter of further interest, we have prepared Diagram I for this 16-year period for Saint Louis, and Diagram J for Rock Island, giving the mean curve and the year when the highest and lowest water occurred on any day of the year. These show the extreme ranges that have been observed at the same dates in these 16 years, and the included space forms what we have called a zone of river oscillations, which includes all our known observations. The three exceptional high waters of 1844, '51, '58, recorded at Saint Louis, are noted on the dia gram, but not included in the zone. We could not give each year's curve without too great confusion or resorting to colors which photolithography would not reproduce. These diagrams furnish interesting data for comparison and study, and information as to probability of high and low water at any time of the year, useful in planning engineering operations. The daily observations are preserved in permanent records and can easily be procured when desired for any special purpose. Comparison of means of gauge-readings at Saint Louis, Quincy, Rock Island, and Dubuque.-In order to see the effect of the natural conditions at Saint Louis and Rock Island, we have compared the simultaneous |