CORRELATION OF THE PREHISTORIC "FLOOR," 273
This line of shattered Septaria probably represents the bottom
of the former river channel, and is the product of the erosion of
the London clay. At certain spots, marine shells are associated
with this bed, and I at one time thought that it must be marine.
I have come to the conclusion, however, that the explanation
is simply this, namely, that during the submergence of the
ancient surface a certain amount of marine erosion took place
before the deposition of the Tidal Silt (No. 8), so that the inter-
vening beds (Nos. 3 to 7) were completely eroded away at certain
spots, and marine shells lived among the broken Septaria
At one spot on the tidal mud flats I found last August a
fine example of the last lower molar of Elephas primigenius
in situ in grey marsh-clay. Unfortunately the exposure was a
very small one, and, being surrounded on all sides by the modern
sea sand, its relations to the other beds were somewhat obscure.
The only clue obtainable was that it was underneath a layer
of shattered Septaria. As this line is so persistent, I feel little
doubt that this is its right position, and that the marsh clay in
which the mammoth molar was embedded was a fragment of a
late Pleistocene marsh. This was some miles north of the
celebrated mammoth bed of Walton-on-Naze, with which it may
probably be correlated. The mammoth remains also, which
are dredged from such situations as the Dogger Bank, the bed
of the Thames, and Torbay, doubtless come from similar late
Pleistocene marsh beds.
These facts are an illustration of the impossibility of inferring
the date of one of these deposits from its level alone, without
working at the details of the local succession. Coming somewhat
into line with these low level Late Pleistocene deposits, I have
also recently found a Late Pleistocene marsh surface on the
floor of the Lea Valley, in the gravels which pass under the
Alluvium.27
In the previous paper on this subject28 the prehistoric remains
from the Rainwash (No. 4) were spoken of as the Earlier Series,
while these from the Buried Prehistoric Surface were referred
to as the Later Series. In that paper certain differences in the
technique of the flint working in the two series were illustrated.
The chief point in which subsequent work has shown the need
27 S. H. Warren, Nature, December 15 (1910). "Arctic Plants from the valley-gravels of
the River Lea."
28 S. H. Warren, Essex Naturalist, vol. xvi. (1908), p. 46. S