it to lower levels into the valleys. At page 124, Quart. Journ. Geol. Soc. vol. xxiv., I described such a flood, in Sind in the year 1866, when a rainfall of 24 inches occurred in ten hours, laying all the surface of the land under water, and moving very great weights a considerable distance. The ancient river-deposits of the Thames basin associated with this flood-borne detritus are generally covered by an upper bed of well-rolled gravel, not exceeding 10 feet in thickness. This bed is often only 3 or 4 feet thick, and lies on the frequently eroded surface of the stratified brick-earths and clays of the middle part of the Thames Quaternary series. Even when for a short distance this surface-bed contains few pebbles, there is no difficulty in identifying it; for it will soon be charged again with pebbles. This thin upper coating of well-rolled pebbles, enclosed often in a stiff brown clay, but sometimes in sand or loess, extends from the river-banks to near the tops of the hills over all the other parts of the series of gravels, and often over the London clay and chalk; it covers a most extensive area, forming a marked contrast to the middle series of laminated brick-earth, false-bedded sands, and stratified clays, which are more confined to the low levels near the rivers or brooks. This bed of upper clay and pebbles slopes to the lower ground (I estimate on an average) at 1 in 260, and falls therefore about 20 feet in a mile. It is washed over the surface of the land smoothly, removing all abrupt heaps of gravel, and leaving no greater irregularities of the surface in the London basin than occasional low terraces. The smoothness and perfect manner in which the gravel is deposited is shown by the fact that the surface of the London clay or chalk is covered, notwithstanding inequalities of the surface, by a bed only 25 feet thick over such a large district.
Fig. 1.—Map of part of the course of the river Aire, Yorkshire.
Map of the course of the River Aire, from Bingley to Castleford. The letters refer to the lines crossing the river, and are for reference to the longitudinal section (fig. 2) of the River Aire, or rather of a line supposed to be drawn at each point of its course 8 feet above the full-water-mark of the river. This map has been accurately reduced by the pentagraph from the one-inch Ordnance sheets.
The river Aire rises near Malham Tarn, in Yorkshire, in the Car-
