A, Tylor on the Amiens Gravel. 323 
supposed to have been deposited before the excavation of the 
last fifty feet of the Somme valley, which excavation, he con- 
sidered, preceded the deposition of the gravels near the Im- 
perial road, Montiers. 
The character of the surface of the chalk at Montiers has 
been discussed at full length in this paper, and shown to be 
concave at the pits; while it is represented as highly convex at 
Montiers by Mr. Prestwich and Sir C. Lyell. 
In the long section C D (Plate IV, fig. 3,) the St.-Acheul 
gravel, at a height of 140 feet above the sea, is shown to be 
separated from the loess at Longueau, at a height of ninety feet, 
by an escarpment of bare chalk. The tramway (Plate LV, fig. 
4d, passing from the Imperial road to the railway, crosses one 
of the supposed bands of chalk marked by Mr, Prestwich. 
But, instead of chalk, there were nine feet of good gravel and 
loess exposed in this cutting. The La Neuville ballast-pit, 
&xposing ten feet of gravel and loess, is also on the supposed 
outcrop of bare chalk (Plate IV, fig. 2: RS on the Plan). 
The outline of the sections © D and I K would at first sight 
seem to confirm the opinions advanced by Mr. Prestwich, that 
the gravel at the 140-feet level might be of a different age to 
that 50 feet below it, 
_ The loess, also, at Longueau, at the 90-feet level, near C, can 
be traced to La Neuville, and then up to the St.-Acheul pits 
continuously. The railway-cutting in La Neuville for half a 
mile is in loess, with veins of gravel (fig. 12); and this is seen 
to be continuous with the St.-Acheul gravel to the north, by a 
number of old pits, and in the tramway. The surface of the 
kis concave in this part of the La Neuville valley, between 
RS and I K, so that gravel and loess would be retained on it; 
While along the lines C D and IK there isa very steep escarp- 
yent, on which no gravel would lie. This escarpment would 
be swept by the stream of the river Arve and the Somme, flow- 
Acheul and La Neuville is of one period, and that it remained 
Spread over the valley of the Somme where the ground was 
foncave enough to retain it. The absence of gravel. on the 
Stee ents and near the river channels is a proof of 
great floods and rapid currents during the Quaternary eae 
i i er 
seg vant - it was sloped and very much in the state it is at 
(Big, 19). 
‘ae M. Guillom’s section the depth of the chalk below the 
Tails has been accurately determined at two points, where the 
