14 NOTES ON THE GEOLOGICAL SECTION AT CHELMSFORD.
tended, when in their normal condition, to throw down very fine
gravel, sand, and mud. But when floods of unusual magnitude
occurred, coarse gravel would be spread here and there on the finer
sediments, to be in its turn again covered by sand and mud. And
the ways of rivers in flood have recently been so singularly well-
illustrated in the case of the little stream called the Ravensbourne,
which flows, like the Chelmsford rivers, through a Tertiary district,
that it seems worth while to give some account here of the changes
in it resulting from the heavy rains of last November (1894).
For about a mile of its course the Ravensbourne flows through
the Lewisham Recreation Ground, and its banks there have been so
regularly turfed and sloped that the effects of the floods could be
seen with an accuracy that could hardly have been exceeded had
the stream been modified simply with a view of demonstrating their
results. In its normal state the little stream brings down
and deposits nothing but sand and mud. Its breadth averages
about 20 ft., and the alluvial flat through which it runs at the
Recreation Ground has a level from 5 to 6 ft. about the average
height of the water. Here and there, where horseshoe-like bends
occur, a shorter channel has been made, and the peninsula
enclosed by the bend converted into an island. These islands
seldom exceed 20 yds. in length and are of less but variable breadth.
A slight embankment prevents the ground from being flooded,
though the floods which everywhere prevailed in November (1894)
so swelled the stream that the alluvial flat would otherwise have
become covered with water. When the floods subsided, the trimness
of the sloped and turfed banks allowed their effects to be demonstrated
with singular precision. Here and there sand in varying amount
might be seen on the grass as high as the edge of the alluvial flat,
and the tops of the islands. But the most curious alterations
appeared in the channels of the stream where alternative courses
existed. In one case (Diagram A) a very considerable mass of
coarse gravel, covered by sand and vegetable refuse, had been
deposited at the mouth of the old course, and a smaller amount of
sand and vegetable matter could be seen thence downwards, as far
as its junction with the new cut. In another case (B) where the
channels on each side of the island were nearly equal in directness,
the floods seem to have at first preferred as a main channel the
course which afterwards became a back-water. A considerable
mass of gravel might be seen from the upper mouth of this