It is particularly interesting for those British geologists who devote themselves to studying the Thames gravels, to have a statement, as accurate as possible, of the results at which I have arrived in a parallel study, which I have pursued for nearly twenty years in the Somme valley.

On the slopes and in the foot of this valley, a series of deposits bear witness to successive oscillations with phases of erosion and aggradation, episodes which brought the river from a high level to its actual course, running on the deposits of its last aggradation, those which are visible being peaty.

There is every evidence, judging by the remains left by the river on the slopes of its valley, of a series of levels, differing both in height and physical state, as well as in their paleontological and archaeological content. These have been called terraces, though this term is very inexact, for they are generally only steps surviving of ancient scooped out aggradations, which in their turn filled up the modern valley. Each is really complex, and, far from being of one date, is the accumulated result of phases of fluvial accretion at different periods of aggradation, which spread out or partially removed, at the beginning and end of these phases, the contributions of lateral phenomena, usually solifluxion, at the time of the deepening of the river bed.