The processes by which land forms are made and continually changed are well illustrated by many phenomena within the area here described. The general principles underlying these may be conveniently discussed under three general heads: (1) Weathering or the breaking down of rocks, (2) Transportation or the movement of disintegrated rock, generally through the agency of water, and (3) Deposition, or the laying down of mud, sand, and gravel, generally after transportation for a considerable distance.

WEATHERING OR THE BREAKING DOWN OF ROCKS

Solid rocks are wasted away by a variety of processes. Some of these merely break up the ledge or stratum into large pieces, and these again into smaller fragments without changing their nature or constitution, so that in the end each fragment is, in all its properties, identical with the original ledge. Such changes are purely physical and, at most, only disintegrate the rock, but do not decompose it. Other agencies actually decompose the rock, changing its chemical constitution as well' as destroying its integrity. Some rocks are more affected by the purely physical agencies of disintegration, others yield more to the chemical agencies of decomposition. If a rock is composed of two or more kinds of minerals thoroughly intermixed, as in case of granite), the decomposition of the one mineral may cause the entire rock to disintegrate. While the several agencies of weathering may readily be distinguished and discussed separately, the actual process which causes the disintegration of any one rock may be, and usually is, complex.

Some rocks, like salt, are readily dissolved in water. Others, like limestone, are dissolved slowly by ordinary terrestrial waters after a slight chemical change. This process is well illustrated in this area. Its results are seen on an exceptionally large scale in the cave region of Kentucky.

Passages Made by Solution.-One of the striking features of the limestones of this region is the occasional presence of grooves or tubes in the surface of the bed. (See PI. II-B, and Fig. 30). These grooves may be as large as a lead pencil or larger than a man's arm. They wind and curve in serpentine forms and are so long that a single slab of limestone generally shows nothing of how they begin or end. Such grooves started by the slow trickling of water between the closely fitting beds. Where well formed they are always on the under side of limestone beds, because the intervening shales are more impervious than the limestones. The latter are more or less cracked or jointed. Through such openings the water descends, but is stopped by the soft dense shale which does not crack. The water therefore trickles along between the limestone above and the shale below. It can dissolve the former but not the latter. Hence it enlarges its passage upward. As the groove is enlarged, the slow trickling may give way to a flow at a considerable rate.

In a similar way many of the joints are enlarged. It is noticeable in many quarries, especially near the surface, that the blocks which compose any one bed do not fit closely but may be separated a few inches. The edges of the blocks are frequently concave, thus making the passage more or less tubular (Fig. 30).

Caves and Limestone Sinks.-The process above described, when carried further, results in eaves. Most caves have this origin and are therefore in limestones. Such caves are, therefore, not primarily rooms but passages. The stream which is usually present is, therefore, not an accident or a mere incident, but may almost be said to be the cause of the cave. If the limestone were 100 per cent pure it would all dissolve, and the cave stream would carry no mud or sand except such as might fall through the roof. In such a case the cave would be due almost wholly to solution. Ordinary limestone, when dissolved, leaves a residuum of mud which gives to the water a rasping effect. Where the limestone contains flints or chert nodules, these, being insoluble, accumulate as gravel in the bed of the cave stream, thus making its wearing effect much greater.

Limestone sinks, or "sinkholes" are caused by the dissolving action of water descending from the surface through joints or cracks to some kind of passage below. Sometimes this passage is a large cave, but it should not be inferred that a sinkhole always, or even generally, indicates the existence of a cave below. A few typical limestone sinks are found in Burnet Woods Park and the, campus of the University of Cincinnati. Excellent examples are seen near the eastern border of the Covington City Park (Plate. IV-A). Others are found on the upland near Licking River, six or seven miles south of Covington. A small area which is pitted to a remarkable degree with such holes is found in the northern part of section 17, Sycamore Township, about midway between Hazelwood and Sharonville.

Very large sinks are sometimes made by the falling in of the roof of a cave, but this region, containing no caves worthy of the name, does not afford examples of this type of sink. Where the cave or passage below a sink, or the passage leading downward to it, is stopped, lakes or ponds result. This is the case with one of the sinks shown in plate IV-A. Where sinkholes are numerous they receive all the water falling on the surface, and continuous surface streams are not found.

Other Chemical Changes.-Aside from the processes involved in the solution of limestone, the only chemical process much in evidence is oxidation. The oxygen of the air in the presence of the moisture of the earth affects even the limestones. The effect may be seen in any quarry or ledge. The rock is weakened, and its lively blue color is changed to a dull yellow. The substance thus sought by the oxygen is iron, and the yellowish color of the weathered rock is due to iron oxide (iron rust). Slabs taken from near the surface in this area frequently have a blue center surrounded by a yellow border, indicating the presence of iron oxide. There is no more iron in the border than in the center, but it is not noticeable as a coloring matter until fully oxidized. Many of the rocks of the glacial drift contain far more iron than the limestone, and are consequently more subject to oxidation. In many cases their color is completely changed and their coherence entirely lost.

A very important agent in decomposing the minerals of igneous rocks is carbonic acid, a gas which animals exhale from their lungs and which results abundantly from decaying vegetation. The work of this agent is called carbonation and is quite as important as oxidation, but, unlike the latter, it is not apparent to the casual observer. It helps to decompose most of the igneous rocks such as the granite and other foreign bowlders of our glacial drift. The decomposition and disintegration of these yield sand and clay as shown before. Moreover, if carbonic acid gas were not contained in ordinary surface waters and ground water, they could dissolve very little limestone, so little that there would be almost no eaves and but very little limestone soil.

Frost.-Of the agencies which merely break the rock to pieces but do not change its nature, the most prominent in this region is "frost," or, more properly, frozen ground water.

Water, on freezing, expands one-eleventh of its volume. The force of this expansion is about one ton to every square inch. Hence any crack in rock, if filled with water, will get larger and larger with every freeze. Ultimately the strongest rock must be ruptured if its minute cracks are allowed to fill with water in this climate. The effects of frost are seen at the foot of every steep rock face. Wherever grading for railroads, roads, or other purposes, has made a steep rock face, or wherever an old quarry has been abandoned and its steep face left to the forces of nature, frost has pried loose fragments of the rock, and these have accumulated as a talus at the foot of the cliff. Railroads which run at the foot of steep rocky slopes, like the Chesapeake & Ohio opposite the mouth of the Little Miami, encounter danger from such falling blocks in the freezing season.

The work of freezing water is not limited to steep bluffs, though its effects there are more spectacular than elsewhere. Everywhere, as deep as the "ground freezes," the effect of freezing in cracks and pores is to rupture the rock and to open the way for the agents which act chemically. Nor is the work of frost restricted to fresh and solid rocks. It continues with equal importance in weathered or "rotten" rocks and even in soil.
Heat and Cold .-Another agent which helps to open cracks is the sun's heat. By causing expansion during the day and in summer, associated with contraction at night and in winter, it is an effective agent, but its effects in this region are not so evident as are those of freezing. It co-operates with frost in the wasting of cliffs and the making of talus. Its effects are often seen in the cracking and bulging upward of concrete sidewalks on a hot summer day. Rocks are broken by this agency as a glass dish or jar is broken by hot water, that is, by the expansion or contraction of one part without the remainder, or of the surface without the interior. For this purpose the heating or cooling must be relatively sudden. This agency is more important in deserts where the contrast between day and night temperatures is greater than here, and where heating and cooling are more sudden.

Plants.-The work of roots in helping to break up rocks is manifest at many places on the bluffs where trees grow with little or no soil, their roots entering crevices in the rock, which are slowly widened as the roots grow. It may, however, be said of this agency as of many others, that its more spectacular effects are not its most important work. For every block of rock actually pried off by the expanding root of a tree, a thousand crevices are slightly widened so as to admit more freely the agents of chemical decomposition. Again, the work of tree roots which challenge our notice, is probably small as compared with that of humbler forms of plant life like lichens. Wherever the face of a rock is blotched with green or gray patches, the minute rootlets of lichens are engaged in a work whose aggregate effect greatly exceeds that of tree roots. In still further contrasting the more noticeable effects with those which escape popular notice, it might doubtless be shown that the aggregate of all the mechanical effects of growing plants is small in comparison with the chemical effects due to the products of their decay, especially to carbonic acid.

Sapping.-An important factor in the breaking up of rocks is the indirect effect of the disintegration of their neighbors. Where the edges of nearly horizontal beds are exposed, and one of the lower beds is easily worn or weathered away, the effect is to withdraw support from the stronger rocks above, which then break off in a vertical face. This process is called sapping. It is very important in the weathering and wasting of all the cliffs in this region. All consist of limestone interbedded with shale, the former strong, the latter weak. It has already been shown how frost breaks off blocks of limestone, but this process is greatly helped by the fact that the shale beneath each limestone bed weathers and wastes away continually. Thus the strong bed above is left, overhanging. Ultimately it would break off of its own weight, even without frost. The weathering of the shale itself is partly by alternate freezing and thawing; partly by alternate wetting and drying; partly by dissolving certain constituents. Vegetation and wind also help.

In a large way this principle of sapping is accountable for some of the steepness of the bluffs in the southern half of this area. It has been shown that the upper half of these bluffs consists of stronger formations than the lower half. The latter is largely of shale which wastes rapidly.

All the processes described above which break down rocks, tend to bring them into suitable condition to be transported. In fact, the whole complex process of weathering may be viewed from a physiographic standpoint as merely preparation for transportation. This latter process, like the former, is effected chiefly through the agency of water, though to a less extent by wind, and in a slow way by gravity without the immediate aid of either water or wind.

The word erosion, which has come to be used in a somewhat general sense, may be taken to include all processes whose effect is to pick up and remove surface materials. It is frequently, though not uniformly, used in more restricted senses. Probably the word is now too familiar to be pressed again into a narrow technical sense. As generally understood, it includes corrosion. This is the process by which a stream wears away its bed, often working on fresh rocks, that is, without previous weathering. Erosion also includes the widespread wash by surface waters before concentrating into streams or even into rills. In a very general way it even includes creep (see below), as when mountains are said to be carved and reduced in height by erosion. The terms creep, unconcentrated wash, and corrosion are specific and definite, and should be used instead of the general word "erosion" wherever the. manner of removing material is to be specified.

A quiet process, attracting little attention but omni-present on slopes, is the slow creep of distintegrated rock and soil under the influence of gravity. The steady and uniform pull of this force would not of itself set the mantle rock in motion, but during any slight disturbance or rearrangement of particles a majority of them come to rest a little down hill from their former position. Such rearrangements result from such alternating processes as freezing and thawing, warming and cooling, wetting and drying, or from the creeping of worms and the movement of roots as the trees above sway in the wind. Nothing in the process attracts attention, but its results are so common that a landscape of steep slopes would look strange without them. Among these effects is the tendency of trees to lean down hill, a phenomenon so nearly universal that it is not fully realized until a picture is seen in which the trees along a ravine are made to stand vertically. The ground near the surface, being most subject to such mild agitation, moves most rapidly, carrying with it the base of the trunk, while at greater depths the roots retain almost their original position. Thus a rude measure of the velocity of creep is afforded by the degree to which trees lean down hill.

Since phenomena like this are as widespread as steep slopes, individual examples need not be cited, but they are found in special abundance and beauty on the steep slopes of the Ohio and Miami bluffs and in the ravines which indent them. The character of the surface formations and soil in this region is especially favorable to creep. Sometimes in the winter or spring when the ground is unusually full of water, the movement is so accelerated that large cracks are opened. Pavements, fences, and walls may be badly displaced. Such instances are sometimes loosely called landslides, and they do indeed grade into true landslides in which a portion of the hillside falls suddenly into the valley. The word slump is a less definite term, often applied to such displacements regardless of the rate of movement.

Tile wash by surface waters, before concentrating into separate streams, differs from creep in affecting surface particles only. Like the latter it is intermittent, though its periods of activity differ from those of creep. It is like the latter in its quietness and liability to escape notice, as well as in the greatness of its effects. Water thus moving is not to be thought of as a sheet of uniform thickness, but rather as a great web or net of small rills which are not constant in position, and none of which pursues its course very far without further subdivision or union with others. Even where grass is absent these rills do not cut channels, because their small power is all used in transporting the sediment with which they are loaded.

Among the familiar effects of such unconcentrated wash is the filling on the uphill side of obstacles such as buildings, bowlders, and (in regions where they exist) stone fences. The effect on a plowed field after a rain is often seen in the filling of small depressions, leaving in their place flat surfaces covered with a web-like pattern of rill marks.

Such rills, whose work is classed as unconcentrated wash, bring to the streams a large part of the mud which great streams like the Ohio and Mississippi carry in such enormous quantities. The exact topographic effect of such wash is not easily stated, but without it the shapes of valleys, especially their borders and heads, would differ greatly from the forms which are familiar.

It was seen above that the weathering away of limestone is accomplished largely by dissolving it. Much of this and of many other substances is thus transported to the sea. Most waters from rivers and wells in northern United States are known as "hard." This is because of mineral matter (chiefly calcium carbonate, the substance of limestone) in solution. The Ohio is not a "hard" stream as compared with others. Out of every million pounds of its water passing Cincinnati, 120 pounds are mineral matter in solution. This means that each minute twenty-six tons of dissolved mineral matter pass by Cincinnati on their way to the sea, a total of 13,683,600 tons per year. In a similar way the Miami delivers to the Ohio more than two tons of mineral matter in solution every minute* (See note 6). In proportion to its volume the Miami carries a much larger load in solution than the Ohio.

While on the one hand this process implies a great wasting of limestone and other soluble rocks, it is, on the other hand, supplying the sea with the material from which new limestone is being made. The material now in solution becomes solid again when taken by animals to build their shells. From the remains of these organisms new limestone is made.

A second way in which the waste of rocks is transported to the sea is by carrying it in suspension. In this condition the individual particles of sediment are very small and remain suspended in the water as mud. Each particle is constantly being drawn downward by gravity and would ultimately settle if the water were at rest, but by the constant commotion of the water it is repeatedly carried upward.

Sand-Boils.-From any bridge over the Ohio, when the water is not too low, or better still from the deck of a steamer, the upward currents of the water may readily be seen making the so-called "sand-boils" or "mud-boils." These are nearly circular patches from a few feet to a few yards in diameter within which the water is evidently rising, carrying upward so much sediment that, in contrast with the surrounding water, these "boils" are distinctly brown. Such localized upward currents, and others which cannot be detected by any surface phenomena, may keep a single particle of mud in suspension continuously from Cincinnati to the Gulf of Mexico. More often, however, one particle probably comes to rest many times within that distance, and may rest at some places for years or even centuries.

Contrast of Streams.-Rivers differ greatly in the amount of mud carried in suspension and in the ratio which such load bears to that which is carried in solution. Thus the Ohio which carries 120 parts per million in solution, carries 230 parts per million in suspension. The major part of its load is therefore mud. On the other hand the Miami at Dayton carries 289 parts per million in solution and only ninety-four parts in suspension. Its water is therefore almost two and one-half times as "hard" as Ohio River water, while the latter is almost two and one-half times as muddy as the Miami. Thus the Miami is seen to carry its load chiefly in solution. Even the Ohio is not a muddy stream. As compared with Missouri River at Kansas City the latter is more than three and one-half times as hard and about nine times as muddy. This stream contributes most of the mud to the Upper Mississippi. After the union of this stream with the Ohio at Cairo, Ill., the waters of the two streams may often be distinguished for many miles, each flowing on its own side of the channel, the waters of the Ohio being light yellow as compared with the darker yellow or brown waters of the Mississippi.

All that portion of a river's sediment which is composed of particles too large to be held in suspension, is rolled or pushed along the bottom, or carried forward by a process called saltation, in which the stones bound along very much as a baseball bounds along the ground, touching it at frequent intervals. The whole complex process of carrying material along the bottom has been aptly called by Gilbert, stream traction.

Effect of Velocity.-While, in supporting sediment in suspension, agitation is the chief requisite, the onward velocity of the water at the bottom is of prime importance in transporting sediment on the bed of the stream. It has been found out by experiment that a current of two miles an hour is able to roll or push over its bed fairly rounded stones having an average diameter of one and one-half inches. Since the velocity of streams at the bottom is much less than at the surface, the Ohio and Miami do not at ordinary stages flow two miles an hour at the bottom. In moderate floods, however, their bottom velocities reach that amount, and in great floods it becomes much larger. It is a principle of physics that the energy of a moving body varies as the square of its velocity, hence doubling the velocity causes the water to strike four times as hard. This enables it to move stones whose diameter is four times as large. In volume and weight such stones are therefore 4 x 4 x 4(See note 7), or sixty-four times as great as those which were moved by the water flowing two miles an hour. If the velocity be trebled the water would strike nine times as hard, and the size of the stones capable of being moved would be 9 x 9 x 9 or 729 times(See note 7 ) that of the stones moved at first.

The greatest stream velocities are not found in our large rivers, but in their smaller tributaries like those which traverse the bluffs of this region. These not only have steep profiles but have much smaller depth than the main streams; hence large bowlders which are barely immersed receive the full force of the current. It need not therefore cause surprise to note that from time to time huge bowlders are carried down our smaller streams. As transportation by rolling and pushing is almost restricted to times of flood, and the water is also much muddier at such times, it may fairly be said that streams do nine-tenths of their work in one-tenth of the time. This is a conservative statement. Some streams, especially intermittent ones, may do ninety-nine per cent of their work in one per cent of the time.

The Scouring of ChanneIs.-The material which is urged along the bottom of the Ohio and Miami at ordinary stages is chiefly sand, and that is carried but sparingly. In great floods, however, the underlying sand and gravel may be scoured out to great depths. During the great flood of March-April, 1913, the Whitewater scoured its channel to a great depth, possibly thirty or forty feet at the Big Four railroad bridge one-half mile above its mouth. Such local scouring may be intensified by the presence of piers which hinder lateral cutting and create eddies in the current, but there is other evidence to show that streams frequently make such local excavations during floods. The habits of the Missouri at Omaha have been studied with some care and the conclusion reached that its local scouring frequently extends to great depths, occasionally to more than seventy feet (See note 8). In building bridges over the Mississippi, where forty to fifty feet of gravel intervene between the river's bed and the limestone below, the surface of the latter has, in more than one instance, been found polished and perfectly fresh, instead of rough and weathered as it should be if the gravel rested on it unmoved for many centuries.

Opposite Cincinnati there are thirty to fifty feet of sand and gravel between the river's bed and the solid rock beneath. Observations have not been made with reference to scour, but it is probable that much of this material is periodically scoured out and carried some distance down stream. The bed rock surface is not everywhere equally deep beneath the Ohio channel. Wherever it lies relatively near the surface it is probably worn down at intervals when the sand and gravel are locally stripped away. How deep this scouring extends it is not now possible to say, but it is possible that the deepening of the rock trough is thus going on at intervals throughout that portion of the Ohio which is included in the area here discussed.

The Miami throughout this area runs several hundred feet above bed rock. Its ability to scour out sand and gravel was demonstrated in the flood of 1913 when it cut out gorges twenty feet deep on its flood plain outside the proper limits of its channel.

Corrasion is the technical name for the wearing effect which a stream has on its channel. In so far as it affects hard rocks, it is due largely to the rasping effect of the mud and sand carried by the stream. If a stream has more power than necessary to transport its load, all will move forward with the water; the mud in suspension, at the same rate as the water; the sand and stones on the bottom at a slower rate. Both will corrode the bottom, loosening new material and thus increasing the load to be carried. Thus an excess of power not needed to transport the load, is spent in corrosion. Looked at in another way, the stream always tends to load itself up to the limit of its carrying power. The chief cutting by such streams is on the bottom of the channel. They are said to degrade their channels; that is, they are degrading streams. The effect of deepening is to make valleys steep and relatively narrow (Fig. 31). Excess of power and steepness of sides are characteristic of the smaller valleys in the bluffs and to a large extent throughout the uplands of this region.

As corrosion embraces all the wearing effects of a stream on its channel, it includes that done on the banks. Such lateral cutting does not imply that a stream has power to spare after transporting its entire load. On the contrary, while it is observed to some extent in streams having excess power, it is most noticeable in streams whose load is so great as to prevent down-cutting. Such streams are apt to build deposits which turn the current somewhat against the banks. The effect of cutting away the banks is to embarrass the stream with a still greater load and to require more deposit. Thus a stream engaged in lateral corrosion works its drift over and over, picking up and laying down the same material many times before its final deposition in a delta or elsewhere. In some streams the picking up process is in excess; in others the laying down process.

This process of lateral corrosion is intimately associated with meandering. All bluffs which rise from the bottom lands are made by this process. Their steepness is due to cutting at the base when the channel of the stream was there.

In making the features which characterize this region, the depositional work of the streams is almost as important as their erosional work. Deposition is always associated with some loss of power. This may result from an actual loss of a part of the stream's water, as illustrated by many streams in the and west which lose volume by excessive evaporation. Loss of power within this area is almost always due to loss of velocity, which may be brought about in many ways. Streams which deposit in their channels more material than they remove are said to aggrade their channels, and are spoken of as aggrading streams.

Alluvial Fans.- One of the most common conditions under which a stream loses velocity is a change of slope. In descending a steep slope the stream uses power not needed for transportation,. in corroding its channel and thus adding to its load. When it emerges upon a gentler slope its velocity and power are in part lost, and it must deposit a part of its load. This loss of velocity is due in part to loss of slope, and in part to the fact that it spreads out after emerging from the deep narrow channel which it cut where its power was excessive. In either case power is lost and the stream drops material in its channel. When the channel has been partly or wholly filled, the stream is still further spread, with a corresponding further loss of power. The result of aggrading its channel above the adjacent land is that the stream breaks over and takes a new course or subdivides, until, in the course of time, it has flowed over and aggraded the slope in all possible directions from the mouth of the ravine. The topographic form thus constructed is called the alluvial fan.

The conditions here described are met in hundreds of short narrow gully-like valleys throughout the area considered. The circumstances are especially typical in the ravines which indent the Ohio and Miami bluffs. The little and often temporary streams in such ravines have abundant power, and thus cut the narrow valleys whose sides frequently increase in steepness from the top down to the stream. On emerging from the bluff upon the gently sloping flood plain, material is dropped in the form of alluvial fans or sectors of low cones sloping in all available directions from the apex in the mouth of the ravine where deposition begins.

Sometimes such fans or cones are found singly, and are so well defined that the material of any one structure may be shown to be approximately equal in volume to that of the gully or ravine from which it was eroded. Often, however, such gullies are so close together. that their attendant fans have not room to develop independently. When they are closely crowded, their lateral slopes may be lost, but the inclination outward from the hills is preserved, forming an inclined plane with a slope intermediate in steepness between the abrupt bluffs on one side and the flat flood-plain on the other. Such a feature has appropriately been called an alluvial slope.

Deltas.-Deposition also occurs where streams discharge into bodies of standing water or into more slowly flowing streams. The topographic form thus produced is the delta. This is not exemplified in a large way within the area studied. Temporary deltas of perfect form are sometimes made in the quieter parts of the large streams.

Probably the most abundant of all stream deposits are those made in the relatively quiet water on the bottom and banks on the inner or concave side of meander curves. These are made at all times and are independent of overflow. The rate of their making is, however, greatly accelerated by moderate floods such as approximately fill the channel. The material laid down is that which is being urged along the bottom. It consists in general of sand and gravel. These are built into a shoal, sloping streamward. Successive additions are put down in rude layers or strata whose dip is that of the surface slope (Fig. 35). The height of these deposits is that of the water surface so long as the stream remains within its banks. As the channel is shifted by lateral corrosion toward the convex side of the curve, this deposit of sand and gravel is broadened on the opposite side.

In the way here described, the entire area over which the channel is shifted comes to be covered with a deposit whose origin is independent of overflow. In the case of most streams which have floods, this deposit is covered by successive layers of mud, laid down during overflow as described above.

In the two cases considered above, the change of carrying power is between one place and another and deposition is correspondingly local. During the subsidence of a flood, power is lost throughout the entire stream and deposition is correspondingly general. In the one case the variation is from place to place and the deposit is local; in the other case the variation is from time to time and the deposit is temporary.

Deposits from Suspension. - Deposition during subsidence of floods involves both material carried in suspension and sediments pushed along the bottom. The former is deposited as a fairly uniform sheet of mud over the area overflowed, but not within the channel into which the diminished stream retires. The great flood-plains in this region are in part composed of mud or silt laid down in this way. This is, of course, especially true of their superficial layers. At almost any place where great rivers are cutting into their banks a passenger on the deck of a steamer may easily see that the upper five to ten feet of alluvium is prevailingly silt in contrast with the lower layers which consist more largely of sand and gravel.

Deposits at Bottom of Channels.-When the flood subsides and the velocity is reduced, much of the material moved along the bottom comes to rest. Sedimentation of this kind affects not only the regular channel but the temporary channels through which the river may flow over its alluvial plain. Sheets of mud deposited from suspension are, therefore often traversed by a network of gravel bands. Where the flood plains are built mainly by overflow, and therefore of silt, such webs of gravel may be found at all elevations throughout the deposit, even though it be hundreds of feet thick. The deposit of gravel in this manner was strikingly illustrated by the Miami flood in 1913.

It has been pointed out above that the Ohio during flood may scour its channel to great depths. The depth of deposit at such places during subsidence is correspondingly great. The fact, therefore, that a stream at ordinary stages runs over a deposit of sand and gravel, is not to be interpreted as showing that the stream is aggrading, or that it is failing to deepen its trough by corrosion.

In considering the origin of valleys, two widely different and sharply contrasted sets of conditions present themselves. The first is that of an accumulated volume of water seeking an outlet to the sea by the lowest possible route. Such a route would be determined by selection from the source downward, and the valley which would later be eroded along that route would be the natural consequence of slopes which were there before the stream, and were simply found by the stream while hunting the lowest possible course. Such valleys and streams are, therefore, spoken of as consequent. The Niagara and St. Lawrence rivers which simply followed up the retreating glacier as it slowly melted back toward the north, are good examples. Parts of the Missouri and Ohio rivers, which were crowded southward to their present positions by the continental ice cap, and parts of the Mississippi displaced by the same means, are of the same class. The Potomac and James, which once entered the sea at Washington and Richmond, simply extended their courses seaward as the continent lifted, and the marginal sea bottom became the coastal plain. In doing so they simply followed the slopes which they found, and are, therefore, in their lower courses, consequent streams.

Origin.- In contrast with streams and valleys of the above type are those which begin at the lower end and grow headward. Streams and valleys of this kind are subsequent (see note 9). They start as gullies. In this case the initial condition of drainage is one of unconcentrated wash. Somewhere on the slope so washed (it may be near the foot), the concentration of water becomes sufficient to cut out a gully. This gully at first carries nothing but storm waters and hence has an intermittent stream. Repeated rains, however, cut it down to the surface of ground water, and when its channel indents that surface there is constant flow.

The contrast between consequent and subsequent streams may then be expressed as follows: (1) Consequent streams find their courses laid out for them by pre-existing slopes; subsequent streams are in a measure independent of such slopes. (2) The courses of consequent streams develop from source to mouth; those of subsequent streams from mouth to source. All of the ravines and nearly all the smaller streams in this area are subsequent.

Down-Cutting.-Down-cutting may go on rapidly for a time, but its effect is to diminish the slope from any one point to the point of discharge. This slope cannot be reduced beyond a certain degree, for the material constantly washing into the gully must also be washed out. If the washing out process fails to keep up with the washing in process, the gully will fill up, and the slope toward its mouth will again be increased. So long as down-cutting is active the side slopes are steep, and may increase in steepness toward the axis (Fig. 31).

Widening .-Since the effect of cutting down the axis is to increase the steepness of the sides, the washing and wasting of these is thus favored and the gully is thereby widened. The deeper and wider the gully becomes the longer are its slopes, and the more material is washed from them into the channel. In the early stages, while these side slopes are short, the deepening process has the advantage and the gorge is accordingly steep. Later the wasting of the sides overtakes, as it were, the corroding of the axis, and there may come a time when the power of the stream in the axis is all used in transporting the sediment furnished to it by the wash of its sides and head. The resulting inability to cut down the channel is shown in the slopes of the sides which become constantly more gentle. If the new valley is alone in a considerable area it widens indefinitely, but if it has neighbors its own slopes and those of its neighbors ultimately meet, and form sharp divides. The slopes of each can then continue to cut lower, but neither valley can widen except at the expense of the other.

Headward Growth .- For reasons similar to those which cause a valley to widen, it is also elongated headward. The headward growth of gullies is, or should be, familiar to all. If the valley cut down without widening, the slope of the sides would have to be vertical, which soon becomes impossible. So also, if the valley cut down without lengthening, the slope at the head would have to be vertical, which is equally impossible. The essential fact in both cases is that the cutting down of the foot of a slope steepens it and thus promotes erosion. Only, this erosion at the gully head is more rapid than that on its sides because more water comes in at the head.

The headward growth of a valley is subject to the same limitations as its widening. So long as it exists along, such headward growth may go on indefinitely, though at a constantly diminishing rate. Sooner or later, however, it meets other valleys leading in the opposite direction and a sharp divide results. The two valleys or streams are then in headwater opposition. Both slopes may then be cut lower, but neither valley can elogate headward except at the expense of the other.

It will be observed that what is here said about deepening and widening is almost equally applicable to consequent streams. Headward lengthening, however, is a distinctive characteristic of subsequent streams.

Branching.-For various reasons the headward growth of a valley is not generally confined to a single line. More than the average supply of water may enter from several directions. This may be caused by initial slopes which, although far too gentle to constitute stream valleys, are sufficient to determine the direction in which gully heads shall grow. Instead of a greater supply of water from one direction, the rock may be weaker along certain lines and thus favor the headward growth of gullies. This is true where the rock is cut by jointing cracks. Such joints are enlarged by weathering and, even though filled with disintegrated rock, erosion is tempted to follow such lines.

Where strata are inclined and outcrop in bands as in figure 23, it is almost certain that subsequent streams and valleys will develop on the outcrop of the weaker beds. Thus it happens that subsequent streams may show a striking parallelism. Where there is no discernible condition governing the headward growth of valleys they branch in all possible directions, and are called dendritic, from their resemblance to twigs on a tree. Subsequent valleys are rarely found simple and solitary. If given sufficient time they increase in number, length, width, and complexity, until the original land surface has disappeared and the entire area has become valley slopes.

>Cross Sections of Valleys

Repeated allusion has been made above to the effects of certain processes on the steepness of slopes. These find their expression in the cross section and profile of the valley. In general, the slope of the cross section depends on the relative vigor of two processes, corrosion of the channel and wasting of the slopes by weathering and unconcentrated wash. The former tends to make the valley relatively narrow and deep. This is the case with young valleys, which are frequently V-shaped. The latter tends to make the valley wide and relatively shallow. This shape characterizes old valleys. If the vigor of down-cutting be great as compared with widening, the side slopes often increase in steepness as the central channel is approached, that is, they are convex upward (Fig. 31). With less proportionate vigor of down-cutting, this convexity disappears. With still less vigor of cutting, the side slopes become compound curves, steepest at intermediate heights and flattening as the axis is approached, that is, the valleys become U-shaped. It is not to be understood that streams in U-shaped valleys are as a rule aggrading their channels, nor even that they have ceased to degrade. In a later stage, valleys have flat bottoms and bluffs. The development of these features depends on meandering, which must first be explained.

Cause.- At any fortuitous curve or turn, the stream's power is to some extent concentrated on the outer or convex side of the channel leaving the water on the concave side with less than its average velocity. At such a time, if the average down-cutting power of the stream be sufficiently small in proportion to its load, its power on the inner side of the curve becomes actually deficient, and deposition takes place in the channel against the inner bank. The effect of this is to narrow the stream, whose power is then still further concentrated against the outer bank. The shifting of the channel which is initiated in this way tends to make curves of somewhat uniform sharpness, their radii depending on the volume and power of the stream and the nature of its load. When well developed, such curves are called meanders or oxbows (Fig. 34). It is to be observed that the conditions of meandering are not limited to aggrading streams, or even excluded from degrading streams. The one essential is that in going around a curve the distribution of power shall be such that the power on the inner side of the curve falls below what is necessary to carry the load; that is, the load must be sufficiently near the stream's capacity so that it shall be locally in excess in the relatively quiet water at the inside of a curve. The Little Miami has good meanders in its lower course, yet it is not an aggrading stream. It has probably ceased to deepen its valley with reference to the adjacent hilltops, which are themselves beginning to cut down, but the valley bottom is likewise slowly cutting down, or at least not building up.

Change in Form of Curves.-The manner in which meanders originate implies a constant change of form. The more the stream cuts to one side the more sharply curved does it become, and the sharper its curvature the more does the stream cut on the outside of the curve and deposit on the inside. The ultimate effect of this tendency is to produce closed curves, formed by the stream's intersecting itself (Figs. 33 and 34). The completed circle or closed curve is then abandoned, and the stream is temporarily straighter than before. In this way, cut-off or ox-bow lakes are formed. For a time such abandoned arcs are in free communication with the new channel, but as the muddy water of the stream enters the ends of the abandoned curve its motion is lost and some of its load drops. Thus a dam is soon built at each end, and the remnant of the old channel becomes a closed basin.

Change in Position of Curves.- If any one meander curve be mapped in successive years, it will be seen to change its position as well as its form. Each meander is found to be moving down stream (Fig. 34). This is because a stream, in flowing transversely across its flood plain, cuts more rapidly on that bank which is on the down-valley side. This down-stream migration of meanders is important. The river at each meander swings only toward its outside curve; therefore, so long as a meander does not move down stream there can be no return of the stream toward the other bluff except by a cut-off. But as a curve which is cutting toward the right moves forward, that is, down stream, a curve which is cutting toward the left occupies its place and the stream begins to swing toward the other bluff without a cut-off. Thus the stream at any one place is observed to swing from side to side, a thing which it would not do if the meanders did not migrate down stream.

Planation.-The work done by a stream in the process of meandering is both destructional and constructional. Its destructional work is done against the outer bank of each curve. This bank may consist of the river's own alluvium, or of the original material in which the valley is cut. In either case the river planes the country to its own level and produces a flat which is co-extensive with the area over which the stream has meandered. This process is planation, and is the essential process concerned in making a flat bottom in a valley which is not being aggraded. Even in such a case there is a coating of alluvium but its thickness is roughly uniform and it is, therefore, not the primary cause of the flat.
The flat bottom is limited by slopes which are commonly steeper than any slopes of the original surface, because made by the cutting of the river at their base. These slopes are the bluffs. Their distinguishing characteristic is their steepness, and their chief significance in the history of the surface lies in the fact that they mark the lateral limits of the stream's meandering and planation. They are, therefore, composed essentially of the material in which the valley is cut. Usually this is not alluvium, but where a stream has cut down beneath its flood plain and formed terraces, bluffs are cut in the older and higher alluvium. In this case those cut in the original country rock may be distinguished as the outer bluffs.

Where a stream, meandering on its flat, flows at the foot of its bluffs, it is widening its trough and flood plain. This is seen most strikingly where the hill rises directly from the river's edge with a slope so steep as to be bare of vegetation. All such cases occur on the outside of the river's curve and indicate that the stream is planing off tile hills which limit its trench. The bluff is, in fact, only the limit of such planation. A good illustration of this process is seen at the "Blue Banks" opposite New Baltimore, and another at the "Cedar Cliff " on the Little Miami just below Miamiville. Where a railroad runs between the river and the foot of its bluff, as the Norfolk & Western does at places opposite Milford and Terrace Park, or as the Pennsylvania does at various places in the Little Miami Valley, expensive work and frequent repairs are often necessary to prevent the roadbed from being washed away by the river in its efforts to widen its valley.

Alluviation.-The constructional work involved in meandering consists largely in the deposit of sediments on the inside of meander curves, the reason for which has been explained above. The effect of this is to follow up the shifting stream with a deposit (generally sand and gravel) laid in oblique layers (Fig. 35) as high as the surface of the river. The bottom of this deposit is the planation surface made in the stream's meandering. The alluvium therefore constitutes a mere veneer whose surface is approximately parallel to the bed on which it rests.

Alluviation also includes the deposition of material from suspension in flood waters. Since the deposition from suspension is due to loss of agitation, it occurs most abundantly where there is a check in the velocity of the water. This is chiefly at, or just outside of the natural banks of the stream, for the contrast between the violent swirl of the flood within the channel, and the conditions over the nearby plain, is greater than the contrast within a similar distance at any other place subject to overflow. If the stream is entirely unable to cut down its channel, this extra accumulation on its banks will result in "natural levees," or a slope of the flood plain away from the stream (Fig. 36). This is very conspicuous on the Lower Mississippi, but is not the case in this area. Along streams which, in the long run, are able to cut down their channels, this extra deposit of silt near the bank is washed into the stream between floods and is therefore not cumulative.

Flood Plain Making.-The making of flood plains by meandering alone involves two very characteristic processes, planation and deposition within curves. The latter may be distinguished as lateral accretion, and contrasted with vertical accretion, which consists in raising the height of flood plains by layers of mud dropped from suspension in times of overflow. In the building of most flood plains, lateral accretion is far the more important process. Some flood plains are built entirely without overflow.

In a section through an alluvial plain the amount of accretion due to each of the above methods may be roughly determined by the nature of the material. Most of the sand and gravel has been deposited by lateral accretion and is not due to overflow, though some has been laid down in channels traversing the flood plain and used in time of flood only. On the other hand most of the silt which is incorporated in the flood plain has been laid down from suspension in times of overflow(See note 10).

Slopes of Flood Plains.-In speaking of planation no account has been taken of the fact that the stream may, while shifting laterally, also be cutting its channel lower or may even be aggrading. In the former case, the surface of planation will not be horizontal but will decline toward the stream. This will become clear by assuming that in figure 35 the channel becomes progressively lower while shifting from c to c'. The steepness of the slope toward the present channel will be determined by the relative rates of the lateral shifting and down-ward corrosion. Since the thickness of the alluvium laid down is approximately uniform, a flood plain surface thus formed has about the same slope as that of the planation surface. Even though vertical accretion be large in amount, and though it be much greater near the stream than at a distance, it is generally insufficient to reverse the slope of the flood plain of a cutting stream, because the thicker layer of mud near the stream is in a relatively exposed position, and liable to wash back into the stream. It may, therefore, be stated as a general rule that flood plains of degrading streams slope toward their streams (Fig 37). This is true of all flood plains in the area here described.

Aggrading streams, on the other hand, habitually build up their channels and are necessarily subject to floods, thus building up their flood plains as shown in figure 36, above. It may therefore be stated as a general rule that the flood plains of aggrading streams slope from the streams toward the bluffs. This is nowhere better illustrated than on the flood plain of the lower Mississippi, where the slope from the river's bank is frequently seven feet in the first mile, and sometimes as much as twenty feet in all (See note 11)

Alluvial Terraces.- Many valleys having bluffs and flood plains have also alluvial terraces, that is, nearly level tracts of land at higher levels than the flood plain. Generally these lie between the flood plain and the bluff though they may be entirely surrounded by flood plain. Frequently there are terraces of different levels, rising like a gentle stairway toward the bluff. Alluvial terraces are remnants of old flood plains. They may be developed in the normal process of a stream's down-cutting, or they may represent some decided change in the stream's life. The latter is the case with the fine terraces along the Miami, Little Miami, and Ohio. In these cases the terraces were made by first filling the great valleys to a level even with the terrace tops. This had to be done by aggrading streams. Later, by an increase of power or a decrease of load, the streams began to degrade and carry away the material with which they had once filled their valleys. Thus they cut to lower levels, leaving the present terraces as the only remnants of the original filling. When made in this manner, terraces on opposite sides of the valley are, of course, equal in height. The business part of Cincinnati and Covington stands on such terraces.

The making of alluvial terraces in the normal course of a stream's down-cutting is illustrated by figure 37. It has been shown that a stream which is cutting downward, while shifting laterally, makes a plain which slopes toward the stream. It is frequently the case with very wide flood plains that their outer margins are too high to be flooded. A stream which has shifted eastward and then again westward while corroding its channel downward, may easily find, before reaching its original position, that the remnant of its old flood plain on the west is too high to be flooded (Fig. 37). In repeated shiftings it may leave any number of flood plain remnants at different heights. This condition calls for no sudden changes in the habits of the stream. In this case, terraces on opposite sides of the valley do not have the same height. All of the younger terraces of this region were made in this way, that is, all the terraces which are intermediate in height between the present flood plains and the old valley filling.

By the profile of a stream is meant the form of its bed considered with reference to its slope only. It therefore ignores the lateral turns or horizontal plan. It may therefore be represented as a curve in a vertical plane (Fig. 38). Corrasion and deposition tend to give continuity to this curve, that is, to do away with repeated changes from gentle slope to steep, and the reverse. The process by which this is done is called grading. On slopes which are too steep the stream has power to spare and will use this power in cutting down. It cuts more near the upper end than the lower end of such a slope. Thus the steep slopes lose part of their steepness. When the stream reaches the flatter slope it may find itself overloaded and obliged to deposit some of its load. This it does at the upper end of the gentle slope, thus increasing its steepness. These processes continue until the profile is harmonious throughout. The final result is not a uniform slope but it is a continuous slope, which is concave upward, being steeper near the headwaters. This is because the stream needs a greater slope where it is small and its power is, to a large extent, used up in friction on the bed. When a stream, even though flowing over strong and weak rocks, has produced such a profile, it is said to be graded.

The ratio of power to load is an important consideration in grading. On the steep slopes the former is in excess; on gentle slopes the latter. The equalizing of slopes, as described above, tends toward a condition in which the power will everywhere be just sufficient to transport the load, but not to corrade the channel, except as the neighboring slopes and hills are simultaneously worn down. It may be shown that the continuous and harmonious slope is reached only when the stream's power has just become equal to its needs for transportation. For, so long as a stream has surplus power it will make a distinction between strong and weak rocks, having its profile steeper on the strong than on the weak; but when power and load have been equalized no such distinction appears.

The small streams emerging from the bluffs in this area are in general above grade. Their surplus power is revealed by (1) the steepness of the side slopes of their V-shaped valleys, (2) the steepness of their profiles, especially near the valley heads, and (3) the inequalities of their profiles, many of which show rapids and small falls of the Niagara type (PI. II-A). Many of these streams are well graded in their lower portions but not near their heads.

The combined effect of all the erosional processes described above is a progressive change in topography. The nature of these changes is so well known that if the form and materials of the original land be given, the succession of topographic features can be predicted until the entire area shall have been cut down so near to sea level that, running water can have no further effect. Regardless of initial form, this is the fate of all land masses unless the process is interrupted. Or, if the topographic features at any stage be given, former features may be known and even, to a considerable degree, the form of the original uneroded surface.

The succession of features of interest in this area is that which begins with a somewhat elevated plain or low plateau more or less abruptly limited by steep edges, such as are represented by the present bluffs.

The first effect of running water on such a surface is the making of gullies at its edge. (See Fig. 32-a). These grow headward and send out branches until, by the work of many such systems, the edge of the area has lost its flatness and consists of hills and ridges. This is the condition in which this area is found for some miles back from the main streams. It does not afford a perfectly exact and simple illustration of the beginning of dissection at the edge of a plateau, for, as noted elsewhere, the present valleys are not being elongated headward in a plateau which is entirely devoid of valleys. The present streams are rather to be considered as deepening the lower courses of shallow valleys already in existence, as will be seen later in the discussion of the history of the present surface. Nevertheless the topography shown at the present time differs little from that of a low plateau edge being dissected for the first time. The largest approximately level tract is found along the line of the Cincinnati, Lebanon & Northern Railway. Its width is gradually being reduced by headward growth of small streams on both sides; on the west by tributaries of Mill Creek; on the east by those of the Little Miami.

So long as the dissection of a plain surface is not complete, many hills and divides continue to rise to about the same height, which is approximately that of the original surface (Fig. 39). They have varying widths, and all are being narrowed by the wasting of their sides, but they cannot be much lowered so long as any part of their flat tops remains. The narrowest will first be reduced to points or crests, after which the continued wasting of their sides will lower their crests. Ultimately every ridge must thus begin to cut down. When practically all of the original flat upland has thus disappeared the country is said to be completely or maturely dissected.

Most of the uplands of this area are in this condition of mature dissection. The hills and divides have been narrowed to points or ridges, but so many of them still rise to about the same height as to leave no doubt that the level thus marked is essentially that of the former plain.

Since the hilltops have held their own up to the time of mature dissection, while the valleys have been deepening, the relief is greater now than ever before. It is also greater than it will be in the future, for it may be shown that henceforth the hilltops will cut down more rapidly than the valley bottoms. (Compare the diagram, Fig. 39.) The stage of mature dissection is therefore also the stage of maximum relief. The slopes are longer than they have ever been before and steeper than they will be at any later stage. The run-off is therefore at a maximum and likewise the power of the streams. Therefore, also, the load of sediment is greater than at any earlier or later stage. Maturity is therefore characterized by maximum dissection, maximum relief and maximum slopes, run-off, stream power, and load. It does not follow that the stage characterized by these maxima should coincide with that at which streams reach grade. Graded stream valleys are sometimes called "mature." If that word is used in both these senses it often becomes necessary to speak of a region as mature while the valleys which dissect it are still young. Such is the case with the mature portions of the area here treated. The confusion thus threatened can perhaps best be avoided by dropping the use of the word "mature" with reference to streams and valleys. For this purpose the word "graded" expresses all that is intended.

After maturity is passed the hills lose their uniformity of height, for the crests of the narrower ones not only begin to be lowered sooner, but are cut down faster than the broader ones. With continued erosion toward base level the relief decreases, and with it the inequalities among hills, until the entire area is but little above sea level. Being "almost a plain" it is then called a peneplain.

The Peneplain in This Area.-There is no simple illustration of this stage of erosion in the area here considered, that is, no peneplain is now found at the level at which it was made. The geological evidence is however complete that the once nearly flat surface of the upland was produced in this way. The present hilltops and ridgetops which are of nearly uniform height are all that remains in this locality of a well developed but subsequently eroded peneplain. When made, it was but little above sea level. The present deep valleys were, of course, impossible at that time. Whatever streams then existed must have been without cutting power, and meandered widely over the nearly flat land. The making of this peneplain is among the most important events in the history of the present surface.

Cycles.-When an old peneplain is elevated erosion has renewed power, and dissection of the surface is again begun. In many respects this second dissection resembles the first. Both begin with a plane or nearly plane surface and end with the same, passing through a stage of mature dissection and maximum relief. The essential repetition of the same processes and features in similar order has caused the term cycle to be applied to the entire round of events from the beginning of erosional work to its close, or sometimes to the time required for all. A region may thus pass through any number of cycles, according to the number of times it is uplifted. This region illustrates the dissection of a low plateau in its second (or later) cycle. Many other cycles may have been completed before the first one of which record is left. The present cycle is loosely called the second, only because the one which ended with the peneplain (represented by the present hilltops) is the oldest of which a record is left.
Rejuvenated Streams.-It has already been shown that when a peneplain is uplifted the power of its streams is again increased. They are then said to be revived or rejuvenated. The effect of this is to cut down rapidly, thus making a deep and narrow young valley within the shallow and wide open old valley (Fig. 41). The same effect is produced when the mouth of a stream is lowered, or when, for any other reason, the cutting power of the stream is increased. The features due to rejuvenation are well shown in the valley of West Fork Creek. (See Plates V-A and IX.)
An interesting phase of revived streams is seen in the Licking and other Kentucky streams tributary to the Ohio. In this case the streams at the end of the first cycle were wandering over the peneplain in an intricate system of meanders. When the peneplain was uplifted and the streams were revived, these streams cut young gorges 200 to 300 feet deep in the uplifted plain, retaining at the same time their meandering course. These curves are now entrenched meanders. The old flood plain on which the meanders developed is represented by the tops of the present bluffs. A new flood plain is being made by continued meandering and lateral corrosion at the present level of the stream. It is plain that such features cannot be produced in a single cycle of erosion, but can result only from rejuvenation. This principle is beautifully exemplified by Kentucky River and by Licking River south of this area. The broad curve of Licking River within this area near its southern margin is of the same character.

Stream Courses of the Second Cycle.-Unless the land is submerged at the close of the first cycle, or the old valleys are effaced in some other way, the second cycle is not in all respects like the first. if valleys, however shallow, remain after uplift, their courses need not be again determined by the process of headward elongation. To a certain extent the country will have a "ready made drainage system." Nevertheless, if the old surface be uplifted without filling, the streams near the center of the area are not at once rejuvenated. The effects of uplift are first felt near the edge where the streams descend to lower levels and their gradients are steepened. Rejuvenation and down-cutting begin, therefore, in the lower courses of the streams, and as these are cut down the renewed activity advances up stream until it reaches the headwaters. This is the way in which rejuvenation occurred in this area when the old peneplain was uplifted. The effects of rejuvenation have now advanced nearly or quite to the heads of the old valleys.

It thus appears that many of the valleys in this area are in their second cycle, but some came into existence in the glacial epoch, and are, therefore, now cutting down for the first time.

Index
Next Chapter
Figures
Footnotes
Plates
Tables