62 GEOLOGICAL NOTES ON A SUPPOSED EARTHWORK
island dividing the stream into two channels for a short distance.
But present or past islands consisting of high ground surrounded by
alluvium, such as the mound we are considering, are comparatively
rare, and are consequently much more likely to be thought artificial
than stream-surrounded portions of alluvial flat.
It is evident that the Stort at one time, instead of flowing north
and west of this oval mound, as it now does, swept round between
it and the Railway Station in a south-easterly direction, and then
turned sharply round towards the north-west, as shown by the
alluvium. At that time the mound was the south-eastern end of a
promontory of high ground stretching from the north-western end of
the present mound towards the house called Rederick. Then (as
so often happens where a river, making a horseshoe-shaped bend,
almost encircles a tract of land) the high ground between the present
mound and Rederick became thinner and thinner, being eaten
into by the Stort, both above and below its south-easterly bend. At
last the stream broke through the narrow isthmus of high ground
connecting the mound we are considering with the high ground
towards Rederick. It then gradually deserted its old channel
south of the mound for a more direct passage north of it, while its
meanderings, since it took its present course, have reduced the
mound to its present length.
The two best examples that occur to me as illustrating this
tendency of rivers to cut through the isthmuses separating their
channels, on the upper and lower sides of their horseshoe bends,
are on the Severn at and near Shrewsbury. Close to Shrewsbury, on
its north and north-west side, the Severn has entirely deserted an
ancient bend northward, and made itself a more direct channel from
west to east. I am informed that water may still sometimes be
seen in the old channel in times of flood. From the high banks on
both sides of this old channel it forms a very striking feature, except
where the encroachments of the builder prevent a good view of it
being obtained. The other case is four or five miles above
Shrewsbury, at a spot called the Isle. There the Severn still makes
a most circuitous bend northward, though the isthmus dividing its
channels above and below the bend is reduced to extremely small
dimensions (See Fig. 2). And it is obvious that sooner or later
the river will inevitably take a more direct course by cutting through
the isthmus at (A).1
1 The River Waveney makes a similar bend around Bungay.