Many remarkable, and, at first sight, very puzzling features connected with the courses of rivers find an explanation when we study the river history. Thus, looking at the Weald of Kent and Sussex, we see that it consists of comparatively low ground rising to a line of heights east and west along the centre, and surrounded on all sides but the south-east by a wall of Chalk downs. If we considered the subject, we should suppose that the drainage of the country would be towards the south-east, which is open to the sea. Not so. All the rivers flow from the central heights north and south,—go straight for the walls of chalk downs, and cut through the escarpment in deep clefts to flow into the Thames and the Channel. This is explained when we remember that the rivers began to flow when the great curve of strata rose above the sea. Though eroded by the sea during its elevation, yet when it rose above the waters the arch of chalk must have been continuous from what are now North Downs to South. And from the centre line of the great turtle back the streams began to flow north and south, cutting in the course of ages deep channels for themselves. The greater erosion in their higher courses has cut away the mass of chalk from the centre of the Weald, but the rivers still flow in the direction determined when the arch was still entire.

We have a similar state of things in the Isle of Wight. Any one not knowing the geological story, and looking at the geography of the Island, might naturally suppose that there would be a stream flowing from west to east, through the low ground between the two ranges of downs, and finding its way into the sea in Sandown Bay. Instead of this the three rivers of the Island, the two Yars and the Medina, all flow north, and cut through the chalk escarpment of the Central downs, as if an earthquake had made rifts for them to pass, and so find their way into the Solent. The explanation is the same as in the case of the Weald. The rivers began to flow when the Chalk strata were continuous over the centre of the Island; and their course was determined when the east and west anticlinal axis rose above the sea.

We shall notice, however, that the Island rivers start from south of the anticlinal axis. The centre of the Sandown anticline runs just north of Sandown, but the various branches of the Yar and Medina flow from well south of this. The explanation would appear to be that the anticline is almost a monoclinal curve,—that is to say, one slope is steep, the other not far from horizontal. Streams starting from the ridge would flow with much greater force down the northern than the southern side, and would cut back their course much more quickly. Thus they would continually cut into the heads of the southern streams, and turn the water supplying them into their own channels.

In its early history a river cuts out its bed, and carries along pebbles, sand and mud to the sea. The head waters are constantly cutting back, and the slope becoming less steep, till a time comes when the stream in its gently inclined lower course has no more power to excavate, and the finer sediment, which is all that now reaches the lower river, begins to fill up the old channel. And so the alluvium is formed which fills the lower portions of our river valleys.

Beyond this, the great rush of waters from melting snows and ice of the Glacial Period has come to an end. The gentler and diminished streams of a drier age have no power to roll flint stones along and form beds of gravel. Gravel terraces border our river valleys at a higher level than the present streams. Periods alternated during which gravels were laid down by the river, and when the river acquiring more erosive force, by an elevation of the land giving its bed a steeper gradient, or a wetter climate producing a greater rush of water, cut a new channel deeper in the old valley. So our valleys in Southern England are frequently bordered by a succession of gravel terraces, the higher ones being the older, dating from times when the river flowed at a higher level than at present. Such terraces may be seen above the Eastern Yar and its tributary streams. In the centre of the old gravels is the alluvial flat of a later age.

The Island rivers cut out their channels when the land stood at a higher level than at present. The old channels of the lower parts of the rivers are now filled with alluvium, partly brought down by the rivers and partly marine. The channels are cut down considerably below sea level; and by the sinking of the land the sea has flowed in, and the last parts of the river courses are now tidal estuaries. The sea does not cut out estuaries. They are the submerged ends of river valleys.

Some idea may be formed of the antiquity of our Island rivers by observing the depth of the clefts they have cut through the downs at Brading, Newport, and Freshwater. But to this we must add the depth at which the old channels lie below the alluvium. It would be interesting to know the thickness of the alluvium. But it is not often that borings come to be made in river alluvia. However, in the old Spithead forts artesian wells are sunk; and these pass through 70 to 90 feet of recent deposits before entering Eocene strata. Under St. Helen's Fort, at the mouth of Brading Harbour, are 80 feet of recent deposits. The old channel of the Yar, at its mouth, must lie at least at this depth.

Before it passes through the gap in the Chalk downs the Yar has meandered about, and formed the alluvial flat called Morton marshes. These marshes stretch out into the flat known as Sandown Level, which occupies the shore of the bay between Sandown and the Granite Fort. What is the meaning of this extension of the alluvium away from the course of the river out to the sea at Sandown? A glance at it as pictured on a geological map will suggest the answer. We see clearly the alluvia of two streams converging from right and left, and uniting to pass to the sea through Brading Harbour. But the stream to the right has been cut off by the sea encroaching on Sandown Bay: only the last mile of alluvium is left to tell of a river passed away. We must reconstruct the past. We see the Bay covered by land sloping up to east and south east, the lines of downs extending eastward from Dunnose and the Culvers, and an old river flowing northward, and cutting through the chalk at Brading after being joined by a branch from the west. This old river must have been the main stream. For it was a transverse stream, flowing nearly at right angles to the ridge of the anticline; while the Yar comes in as a tributary in the direction of the strike. Of other tributary streams, all from the right are gone by the destruction of the old land. On the left streams would flow in from the combes at Shanklin and Luccombe—streams which have now cut out Shanklin and Luccombe chines.

Passing the gap in the downs the river meandered about, and, with marine deposit, washed in by the tides, formed the expanse of alluvium which occupies what was Brading Harbour,—a harbour which in old times presented at high tide a beautiful spectacle of land-locked water extending up to Brading. Inclosures and drainings have been made from time to time, the upper part near Yarbridge being taken in in the time of Edward I. Further innings were made in the reign of Queen Elizabeth; and Sir Hugh Middleton, who brought the New River to London, made an attempt to enclose the whole, but the sea broke through his embankment. The harbour was finally reclaimed at great cost in 1880, the present embankment enclosing an area of 600 acres.

The history of the Western Yar is similar to that of the Eastern. The main stream must have flowed from land now destroyed by the sea stretching far south of Freshwater Gate. All that is left is its tidal estuary, and the gravel terraces and alluvial flat formed in the last part of its course. Of a tributary stream an interesting relic remains. For more than 2 miles from Chilton Chine through Brook to Compton Grange a bed of river gravel lies at the top of the cliff, marking the course of an old stream, of which coast erosion has made a longitudinal section. This was a tributary of the Yar, when the mammoth left his remains in the gravel at Grange Chine and Freshwater Gate. Down the centre of the gravels lies a strip of alluvium laid down by a stream following the same course in later days. The sea had probably by this time cut into the stream; and it most likely flowed into the sea somewhere west of Brook. In the alluvium hazel nuts and twigs of trees are found at Shippard's Chine near Brook.