Page:Quarterly Journal of the Geological Society of London, vol. 26.djvu/742

540 presumption would be strong that the river flowed in the same direction as the present stream, were there no evidence in the materials of the gravel to show that the river ran through the Lower Greensand and Wealden beds. In considering what extent of land existed to the southward of the present coast-line when these gravels were formed, a comparison with the very similar valley at the eastern end of the island gives indications worth notice. At Freshwater a stream rises close to the beach, at a level below that of high water, and flows northward to the Solent, by Yarmouth. At the other end of the island the Brading river has two branches, one rising at the back of a beach protected by groynes, in Sandown Bay, and flowing for its entire course below the level of high water; and the other draining a considerable area towards Arreton and Niton.

Mr. Godwin-Austen has observed with regard to the Sandham level, which constitutes that part of the valley of the Brading river which is below high-water level, that excavations for deep drains show no signs of estuary deposits, or evidence of the sea having formerly occupied it; and the same observation applies to the Freshwater valley. Not only is there no evidence that the sea has ever occupied these valleys, now not much above half-tide level and artificially protected from the sea, but the succession of gravels down to the sea-level seems to prove that both valleys have been gradually deepened by fluviatile action only.

Both valleys traverse the chalk under similar circumstances; the thickness is the same, and the strata are nearly vertical. Fig. 13 (Pl. XXXVII) shows the transverse sections of the two valleys through the chalk drawn to the same scale, and if the sectional areas may be considered as roughly proportional to the extent of the river-basins, the ancient Yar was at least as important a stream as the Brading river. The latter at present drains 24 square miles, to the south of the chalk range, or $1⁄6$ of the area of the island; and gravels on the cliffs about Sandown show that the branch of the river which now rises near the beach is the representative of a much larger stream, which, when flowing 100 feet higher, drained land then occupying the position of Sandown Bay. It is therefore not an improbable supposition that the gravel with Elephant remains on the cliffs at Brook and Grange was included in the same river-basin as the Freshwater gravel, and that the streams now entering the sea between Blackgang Chine and Compton Bay were tributaries of a river flowing northwards through the chalk range to the Solent. The difference of level in the gravel is no more than is due to the natural fall of the watercourses, and the waste of the cliffs now going on shows that a considerable area of land has but recently been destroyed.

The chines by which the streams at the back of the Isle of Wight enter the sea are probably due to the alteration in the drainage consequent on the destruction of this land. The streams ending in chines present no unusual features until they come within a short