PLEISTOCENE CLASSIFICATIONS. 273
Is it possible that the suggested early glaciation of Hert-
fordshire, etc., from the west or north-west (bringing Bunter
pebbles and Rhaxella chert over the Chalk escarpment) may
belong to this stage ?
STRATIGRAPHICAL EVIDENCES.
With regard to the rival claims of pure palaeontology (repre-
sented by the theory of Hinton and Kennard) (15) and of strati-
graphy, to be the main basis of Pleistocene Classification, I
cannot do better than quote the words of a palaeontologist, Dr.
Gertrude Elles, when lecturing upon Graptolites to this Associa-
tion : " Palaeontology should be the handmaid of stratigraphy,
never its mistress." The lecturer emphasised the fact that
the true succession of the Palaeozoic rocks could not be inferred
from a Palaeontological study of the fossils, but that, in the
first instance, the true succession of the life forms must be learnt
from stratigraphy. That is even more the case with smaller
divisions of time (like the Pleistocene) in which local migrations
play such an important role.
Unfortunately, the stratigraphy of superficial deposits on the
land surface is by no means such a comparatively straightforward
proposition as that of marine deposits. It is full of pitfalls and
dangers, even to the most experienced. Flints are very durable
and easily transferred from an older deposit to a newer one,
and this may sometimes occur with but little appearance of
rolling. The superficial re-distributions occasioned by soil-creep,
by the penetration and subsequent decay of the roots of forest
trees, the work of burrowing animals, the underground solution
of calcareous material and shells, the frost and snow of the
tundra conditions of the Magdalenian and Ponders End Stage,
and the like, are far more universal and penetrate more deeply,
than is commonly recognized. Even a level plain with no
appreciable slope, affords no security against the " trailing "
and festooning of the sub-soil.
It is remarkable how few instances we have of direct contact
between palaeolithic and glacial deposits. And as, in most
cases, such instances are subject to some element of doubt
(even if they support one's own view !) the total cumulative
evidence is by no means so weighty as one could wish. The
following cases seem to me notable :—(16)
Couliege (Jura), and Bohan (Aisne Valley).—At both these
localities Lower Palaeoliths have been recorded from deposits
believed to be later than the Rissian Alpine moraines.
Garonne Valley.—Acheulian implements have been re-
corded by Obermaier from many localities in deposits directly
overlying Penck's Third Terrace of fluvio-glacial gravels. The