yet far from level, there being numerous flat ridges 5 feet or more above the surrounding lands, and in some cases so high as to be barely covered by the highest floods. On the other hand, there are many depressions occupied as the beds of lakes and swamps, and a number of sloughs or other water courses into which the drainage of the land concentrates. As a general rule, the land is highest near the bank of the river and slopes gradually away, about 6 feet in the first mile and more gently from that point on. Consequently, the general tendency of the water draining off after rains, or during overflows, is to concentrate into bayous or creeks, which, running away from the river, eventually unite with larger streams or rivers flowing at or near the foot of the bluffs that border the river bottom. In some cases, however, the smaller creeks or sloughs, instead of draining away from the main river, run back to it after a short course, and thus cause complications in levee construction which will be referred to under that head.

This alluvial bottom land has all, or nearly all, been formed by the river by much the same process as is now going on, and all the material that forms this land having been deposited by the river, it is subject under favorable conditions to be again moved by it.

The materials are generally clays and sands, either pure or mixed together, with occasional deposits of gravel, the latter, however, being usually found in small quantities and in thin strata. The sands, especially the fine samples, and some mixtures of sand and clay, are easily eroded by the current, but there is one variety of stiff, blue clay that resists erosion admirably, though even it will yield to a prolonged and severe attack of the river current. These materials exist usually in strata of different thicknesses, varying from many feet to a fraction of an inch. Sometimes these strata are horizontal, or nearly so, and for long distances are approximately of a uniform thickness, while in other cases the strata run only short distances before thinning out and disappearing.

The river bottom is, as has been said, bordered by bluffs well above overflow. These bluffs are composed of various strata of sands, clays, gravels, and some harder material locally called rock, from comparison with the softer clays. As a rule, the bluffs resist erosion better and are cut away more slowly than the bottom land, but they also, if subjected to a long-continued attack, are eventually washed away. In a few cases there exist hard strata of conglomerated gravel, cemented by an iron salt, that is eroded, if at all, with great slowness. When the strata below this can be

reached, they are washed away, and the hard stratum being undermined, breaks off in fairly large pieces. In a few cases, the river not having gotten underneath this hard stratum, and having washed away the softer sands and clays above it, the hard conglomerate is left as a non-erodable bar in the channel. One such, a few miles below Memphis (230 L), had to be removed by blasting.

In general, therefore, the banks of the river are subject to erosion by the current, and within such banks the river follows its crooked course, eroding and cutting away its banks in one place and depos iting the material in another, and in this manner constantly shifting its channel.

The course of the river, as shown on the maps, consists of a series of bends of greater or less curvature, with an occasional stretch of fairly straight river. The river along the bends is usually narrow, averaging less than half a mile in width at low water, while in the crossings from one bend to the next, and in the straight reaches, it is usually wider. The deep channel usually follows the bend, keeping close to the concaye banks, and in such cases low-water depths of 100 feet or more are found. In crossing from one bend to the next below, and in the straight reaches, the current spreads out, and it is in such localities that the bars are usually found, and on these bars available low-water depths of less than 6 feet have been frequently met with.

The river at high and at low water differs greatly, the difference between extreme stages varying, within the limits of these districts, from 40 to 50 feet. The discharge varies from a minimum of about 70,000 cubic feet per second to a maximum of about 1,750,000 feet, or a ratio of about 1 to 25.

The current is as variable as the other elen.ents involved. Owing to causes to be afterwards described, the banks of the river are far from smooth, while the bottom usually consists of a series of cross waves or ridges, and the water flowing along and over these irregularities of bank and bottom is turned from its direct path into whirls and eddies, and irregularities of all kinds. The general tendency of the current is, of course, to flow downstream in filaments approximately horizontal and parallel to the bank, but the projections of the bank turn it obliquely in a horizontal plane, and the irregularities of the bottom turn it obliquely in a vertical plane, and the whirls and eddies cause it in places to flow upstream, and as a result the actual direction of flow at any point may be down

stream or upstream or across the river, toward the surface or toward the bottom, or in any oblique direction. If any cross section of the river be taken, the flow past it is not regular nor uniform, for not only do the velocity and direction of flow vary greatly at different parts of the cross section, but the total flow past is constantly changing, varying in a few moments from a minimum to a maximum and back in irregular pulsations, and the time of a maximum flow in one part of a cross section may not be the time of the maximum in another, and, in fact, almost all imaginable irregularities are found. When we are considering only the discharge of the river, we, of course, care only for an average of the conditions of flow, and the irregularities themselves are of no value, but in other connections these irregularities are of great importance, as will be seen in the consideration of the subjects of sediment and bank caving. Of these different irregularities the most important are probably what are known on the river as eddies, and the formation of these will now be more fully described.

EDDIES

The banks of the river are very irregular in composition, and it very frequently happens that, as the bank is cut away by erosion, a projecting point will be left that resists the attack of the current better than the bank above. Such a projection of the bank is called a "false point." The current flowing along the bank above this false point is turned by it away from its direction parallel to the bank obliquely toward the center of the stream. (Plate I, Fig. 1.) By friction the water at and below the point is carried off with this current, and its place is taken by water flowing upstream along the bank below the false point, and there is thus formed a whirl or eddy. Such eddies are among the commonest features of the river and may be noticed below almost every false point, or similar locality, and in almost every "pocket" in the bank, and, as we shall see afterwards in discussing the caving of banks, they are of great importance.

If this false point extends well down under the surface as a sort of cross ridge along the less steep subaqueous slope, the deflected current has now a vertical component in its motion, and carries away with it in a partly vertical direction the water below the false ridge, and the place of this is filled by a downward flow below the false ridge through the axis of the eddy, and if the suction be great enough there is formed in this manner a regular vortex, such

as is seen in emptying a bath tub. Such vortices are quite fre quently seen on the river, and where one exists, the velocity of eddy flow is usually much greater than the general velocity of the river current. Eddies have been noticed on the river 800 feet long by 200 feet wide, and with vortices at the center a foot or more across, into which the water tumbled, carrying with it floating logs, etc., of considerable size.

Sometimes, when the river velocity is high, these rotating vortices will be snatched by the action of the current bodily away from the locality where they were formed and will go whirling downstream. The hollows of the vortices thus torn away will quickly fill up and the water from without continuing by its inertia to flow in toward the vortex, there will result an appearance as if the vortex were coming up from below and overflowing. Such an occurrence is called a "boil." When such a vortex has been thus snatched away, a new one soon forms in the same place. This in turn may be snatched away, and so on. These eddy vortices seem to constitute one of the important features of the river flow, and their importance in connection with caving banks and in keeping sediment in suspension is very great.

Besides the eddies just described, which we may call "suction" eddies, there is another kind which is also of some importance. Suppose a condition such as shown on Plate I, Fig. 2, where the main current, after flowing along the side of a sharp point, passes abruptly over against a bank composed of some comparatively resisting material. In such a case there is, of course, a tendency of the current to be drawn sharply around the point, but by its motion the main flow is carried across to the other bank, against which it impinges somewhat directly. In the same manner as the water striking the bow of a boat is thrown up into the air, this current impinging thus sharply against the bank is deflected, and part of it flows upstream along this bank and a reverse current or eddy is formed above the point of impingement. Cases of this kind are occasionally found in the river and these, which we may call "pressure" eddies, differ essentially from the above described suction eddies. In the suction eddies the tendency of the suction is to draw the surrounding water into it, but as it comes it brings with it the effect of its previous motion, and instead of flowing directly into the vortex, it whirls around, and failing to immediately satisfy the suction, that suction tends to increase, and the reaction of this increased suction causes an increase in the velocity of inflow, so

that in a suction eddy the velocity is sometimes very great. Pressure eddies, on the other hand, have no such tendency to an increased velocity, and they usually move relatively slowly, and are of importance, as we shall see later, on account of the deposits made in them. At Memphis an eddy of this character has been observed that was nearly a mile long and 2,000 feet wide. One other point specially noticeable is that the flow in pressure eddies does not seem to be regular, the eddy apparently filling and emptying in almost regular cycles. At Memphis it was noted in 1888 that the large pressure eddy filled and emptied itself in pulsations about once every half hour. At one period of the pulsation the flow in the eddy would be upstream, and at another downstream, while the reading of a gage near the foot of the eddy would vary as much as half a foot with each pulsation. In places where the conditions are such as to form a pressure eddy there will usually be below the point a suction eddy also, and in the locality just described, namely, Memphis (Plate L), we find both a pressure eddy on the Memphis side, causing a deposit, and a suction eddy on the Hopefield side cutting the bank away below the point.

TRANSPORTATION OF MATERIAL

As has been stated, the river is constantly cutting away its banks or bed in one place and building up new formations in another. This transportation of material it accomplishes in three different ways, on its surface, in suspension, or along the bottom, and as these three methods of transportation are essentially different, they will be considered separately.

MATERIAL CARRIED ON THE SURFACE

This consists almost entirely of ice or "drift.”

From the latitude of the part of the river under discussion trouble from ice would not be expected to be and has not been very great. Near Cairo the river does indeed at times freeze over, and there is a tradition that even at Memphis the river was once so frozen that it was possible to walk across it, but such an occurrence is very rare. Of course, during a sudden thaw, especially if accompanied by a rise, a large amount of floating ice is brought down from the upper rivers and tributaries, and this at times gorges in narrow places, but these occurrences also are rare. During the runout of the ice navigation is interfered with, but the delays thus caused are not usually of long duration. In changing the regimen