Appendix B: Ring-Ditches
Ring-ditches (i.e. ditches presenting more or less circular
plans, sometimes interrupted by causeways) appear as
crop-marks in many air photographs; they are difficult
to interpret and have been referred to in the following
Inventory only when they occur near an area of Iron
Age or Romano-British settlement. The full range of
such crop-marks appears on the accompanying plan,
which shows the relationship of all known Prehistoric
and Roman monuments in the area of Upper Thames
river-gravels, extending from Gloucestershire into
Wiltshire. For a summary discussion of the ringditches noted within the area described in the volume
(approximately 150), and problems of their interpretation and dating, see I. F. Smith in Archaeology and the
Landscape (ed. P.J. Fowler, 1972), chapter vi.
Some ring-ditches are likely to belong to ploughedout Bronze Age round barrows, as shown by the excavation of examples in the area of the Chessels
Romano-British settlement at Lower Slaughter (1).
Intensive occupation in the gravel terraces (where most
of the examples have been noted) seems frequently to
have resulted in the obliteration or degradation of
pre-existing earthworks; for instance, linear ditches
are sometimes seen to cross ring-ditches without interruption, indicating that any former mound or bank was
already flattened. The circumstances contrast markedly
with those observed in upland situations on chalk or
limestone, where earlier monuments were usually left
unploughed in Iron Age and Roman times. But there is
an occasional indication that certain ring-ditches still
enclosed mounds in the Iron Age and Roman periods.
At Lechlade (2), one of the ditches flanking a track
shows faintly in comparison with the other where it
intersects a ring-ditch, suggesting that the ditch was
cut through a mound, subsequently flattened, leaving
traces of only that part of the ditch which penetrated the
subsoil. At Kempsford (7) a ring-ditch is seen in a space
relatively free of ditched features and approached by
apparent tracks (Plate 60). The suggestion here is very
strong that a mound still stood at one side of the open
space, although at some time the ditches of rectangular
features impinged on its S.W. side.
Circular or penannular crop-marks with exceptionally broad ditches suggestive of recutting occur at
Lechlade (6) and at Kempsford (6–7). A circular ditch
of this kind in the area of another settlement at Lechlade (5) proved under excavation to be Roman in date
and of uncertain purpose.
Small penannular or semicircular features, about 30 ft.
in diameter and with very narrow ditches, dispersed
among linear ditches at Fairford (5) and Lechlade (8),
and clusters of similar crop-marks at Lechlade (2)
and (6), Kempsford (6) and Great Rissington (2), may
represent the sites of circular houses. The patterns
are comparable with those of unploughed Iron Age
hut-circles at Hod Hill in Dorset (Dorset, III, 263–5).