OF THE THAMES VALLEY.
79
We have already described the deposits filling the channel
exposed in the quarry sections, and have cited Mr. Holmes'
description of those seen at the subsidence, but we have
reserved for this juncture mention of one of the most important
facts brought out by this evidence. Suppose we have a stream 
across any part of the lower course of which a hard resistant
barrier is rising, the strata above the barrier being soft and
easily corraded. If the stream is strong it will cut down the
barrier as it rises, if it is too weak then it cannot. The result in
the latter case is this ; a kind of basin is formed above the
barrier, the water gradually rises and runs over the edge of the
dam, but the basin acts as a filter, and the coarser portions of
the detritus of the stream's bed are deposited in it, and then, as
the current is checked, the finer material also is dropped until
the stream escapes over the dam without any burden of detritus,
and so being deprived of its the tools, cannot corrade at all.
The coarsest detritus will be found nearest the barrier, and will
there attain its thickest development, and if traced higher up the
stream it will be found to gradually pass into or give way to the
finer sediment. Now this state of affairs is precisely what we
find exhibited in the sections exposed in the quarry. As noted in
our description, the gravel is thickest at the southern end of the
channel, i.e., the portion immediately above the resistant barrier
of Chalk. Traced to the north it thins out, giving place to sands
and loams, and the passage between the different kinds of
sediment is singularly complete, every gradation being repre-
sented.13
As time goes on the water in such a natural pond rises
higher and higher until it is in a position to find another outlet
along any easier path which may offer itself in the district. In
this case such an easier path did present itself, viz., in a direction
at right angles to the course then followed by the lower part of
the stream. If the Mardyke turned to the west below Stifford
instead of continuing to flow southwards, then it would escape
the obstacle to the south by passing, not across the fold, but
parallel to its longer axis.
13 In 1905 we examined a new section which had been opened to the north of the large
quarry along the line of the old channel. This section showed loam and clay to a depth of
12 feet, and the base of the clay was not exposed. The clay was well laminated, with a few
pebbles here and there, and from it we obtained a pahaeolithtc flake, and a few land and
freshwater shells. The beds were dipping gently to the south, i.e., in the direction of the
channel.