carry its load of material and the sediment is dropped. The larg and coarser particles are, of course, deposited first near the bank while the finer material is carried on farther into the interior an eventually settles in the swamps and lakes. By this process thbottom land is built up most rapidly near the river and more slow! farther away. As a result of this the banks of the river are usually! the highest parts of the bottom and have a constant tendency to get higher, and would in time probably be built up to at least an average flood stage if it were not for the fact that these high banks are constantly being destroyed by caving. RAPIDITY OF BANK BUILDING The process of bar building goes on, as a rule, more regularly and not as spasmodically as bank destruction, but as the average siz of the river cross section seems to remain about the same, the total volume of deposits must about equal the total volume of bank destroyed, and, again, as opposite the bends mentioned on page 2) as places where long-continued caving had taken place, the river has remained of generally the same width, the bars on the conver side must have grown up about as fast as the concave banks were cut away. Cases of sudden and rapid fills are not quite as numer ous or noticeable as sudden large caves, but there have been several cases in the Plum Point reach work where fills of more than 30 feet have taken place over a considerable area in a single high water, and during the formation of the Memphis sand bar a fill of 45 feet in a single year was noticed. The course of the river consists, as a rule, of a series of bends of greater or less curvature, with short, straight stretches between one bend and the next below, and an occasional straight reach of considerable length. In the bends the current from its centrifugal force impinges quite strongly against the concave bank, and the velocity of the water close to the surface of this bank is quite great. and consequently its erosive force is great, especially at high stages. and, as a rule, such concave banks are rapidly cut away. The convex banks, on the other hand, have the current thrown away from them; the velocity close to the surface of the bank is small, while the water passing over it contains more or less of the sediment cut away from the opposite bank and brought over by whirls and cross currents, and hence on such banks deposits are made, and these banks grow, following in their growth the caving of the concave bank. In such bends, therefore, the cross section of the river con sists of one steep and caving bank and one flat and making or The width of the river in such places varies at low water fr about 1,500 to 2,500 feet, with low-water depth of from 50 to 19 feet, the line of greatest depth being about 300 feet from the caving bank. A bend curving in one direction is usually followed by one that curves in the opposite direction, and as the main current hugs th concave bank it is compelled, in going from one bend to the next, to pass from one side of the river to the other, and there occurs what is known on the river as a "crossing." Crossings are, therefor. formed usually at the foot of a bend. In making a crossing the current having no longer a bank to guide it usually spreads out, and as at such localities the high water cross section is much greater than in the bend, the velocity is somewhat reduced and part of the load of sediment taken up from the caving bank is dropped, and a bar is built up. As the water falls after the flood these bars act to some extent as dams, and through them the low water has to cut its own channel. This action of the river in cutting out the deep bends, building up the bars in the crossings during high water, and the cutting out of the bars and washing the material into the pools below during low water is the regular normal cycle of the river. That this action took place annually has long been known, and in August, 1899, a party was sent out by the Commission to make a study of the details of the manner in which it took place. This study has not been finished, but the observations taken to date will be found in the reports of the Seeretary of the Commission, printed in the reports of the Chief of Engineers (1900, page 4746; 1901, page 128, Supplement). The bars made, as just described, form the chief obstacle to lowwater navigation. They are frequently built up completely across the river to above the water surface at the preceding low-water season and are at times composed of material relatively coarse and difficult to erode. In cutting the low-water channel there is usually nothing to guide the river, and consequently the channel that is cut shifts in place from year to year, and at times the river at tempts to cut two or more such channels at a single crossing, with the result that all of them are deficient in depth. If the fall of water after the flood be slow, regular, and gradual, the channel cut will usually be good, but if it be rapid or be broken by short intermediate rises, the channels are apt to be shallow and unsatisfactory. Though caving usually takes place along the concaved bank of a |