tell from the appearance of the surface that such a cavity had existed.
There must have been a wash of gravel all over the sand to have
furnished the material to fill up such cavities as that shown in
fig. 31; and since that wash of gravel there is no indication of any
subsequent deepening of the pipes in the chalk.
The Thanet sands (b), 25 feet thick at B, at the north of fig. 31, are seen lying in situ on the chalk.
Where the formation of the great pipe A has undermined the Thanet sands, b, the latter have fallen down to the bottom of the pipe A, which is funnel-shaped, and filled up 20 feet without any mixture of gravel whatever. The sands moved from their proper position are indicated by the letters b' b'; the bottom part of the pipe A is not shown in this drawing. Under C, masses of sand, b b, are shown falling into the pit by gravitation, from the removal of the chalk by chemical action. The gravel series c c is 30 feet thick above A, and is similar in mineral character to the covering bed e.
Mr. Prestwich has explained the formation of pipes by chemical agency in the chalk.
Fig. 32 is a drawing of other beak-like projections in the chalk
Fig. 32.—Section near Knighton Mill, Isle of Wight.
where it has been acted upon by denudation previously to the deposition of the gravel near Knighton Mill, Isle of Wight.
Fig. 33 (p. 98) is a section of a fissure or pipe in the highly inclined chalk on Brading Down, Isle of Wight, which is a remarkable contrast to the ordinary pipes which are filled with sand or gravel.
The general conclusions to be drawn from the sections of the Thames-valley fossiliferous deposits appears to be that they are a series of variable sands, loams, and clays intercalated between the upper and lower gravels.
The false-bedded sands are below the loams and clays at Erith, and above them at Grays, while at Crayford they are in the middle of the series. This may simply arise from the principal current changing from one side of the old river to the other during the deposition of the Thames Quaternary series. By this means, where the current became more rapid, sands would be deposited, and where
VOL. XXV.-PART I. H
