406 WARREN UPHAM, ESQ., M.A., F.G.S.A.. ON 
Levelling done for Mr. Alfred Tylor by the railway 
engineer, M. Guillom, gave the altitudes of the bottom 
and the original surface of the large St. Acheul gravel pits 
(two are each a quarter of a mile long), the richest in flint 
hatchets among the numerous excavations in and near 
Amiens, as respectively 140 and 160 feet above the sea, or 
about 75 and 95 feet above the river, which is a half-mile 
distant to the north. The pits are not at the upper limit of 
the gravels, for Tylor remarks, “The sections near Amiens 
show the valley gravel continuous from a height of 200 feet, 
at St. Acheul : « tothe River Somme (coated over by 
a nearly uniform warp of loess), and laid at a low gradient 
not exactly parallel to the surfave of the chalk, but ration in 
its concavities.” These higher gravel beds, however, having 
no excavation, it is not known whether they contain stone 
implements and fossil bones; but generally, according to 
Ladriére, these are absent or rare in the highest beds of the 
valley gravels throughout northern France. In the vicinity 
of Montiers, one to “two miles north-west of Amiens, nine 
gravel pits, containing worked fiints and bones of extinct 
animals nearly as at St. Acheul, are shown by 'Tylor's 
description and map to range in height from 77 to 155 feet 
above the sea, or from 17 to 95 feet above the Somme, the 
upper limit of the excavations being the same in relation to 
the river as near St. Acheul. The distinction of upper and 
lower gravel deposits, which Prestwich and Lyell made 
prominent in their writings, was pronounced by Tylor, as 
it seems to me with sufticient reasons, to be seldom definitely 
observable, the series, where developed at considerable 
heights, being usually continuous thence down nearly or 
quite to the bottomland and river. 
Ladniere, who in 1875 and ensuing years has extensively 
examined and described the Pleistocene valley deposits of 
the Seine, Somme, and other river basins of a wide region 
extending northward into Belgium, divides these deposits 
into three somewhat similar series, successive in age and 
stratigraphic order, each of which cun be traced from the 
bottoms of the valleys up to the plateaus, though not to 
their greatest heights. These series, each consisting of a 
regular sequence of gravel, sand, loam or loess, etc. 
representing three distinct stages of the Pleistocene period, 
are doubtless to be correlated with stages of advancing 
glaciation, interrupted by times of decrease and recession 
of the European ice-sheet. The researches of Ladriére thus 
