ers
only a part of the field (_Proc. Soc. Antiq. Lond._, xxiii. 120).]

[Illustration: FIG. 19. 'DRAGON-BROOCHES' FOUND AT CORBRIDGE (1/1). (P.
44.)]

The contrast between this Romano-British civilization and the native
culture which preceded it can readily be seen if we compare for a moment
a Celtic village and a Romano-British village. Examples of each have
been excavated in the south-west of England, hardly thirty miles apart.
The Celtic village is close to Glastonbury in Somerset. Of itself it is
a small, poor place--just a group of pile dwellings rising out of a
marsh, or (as it may then have been) a lake, and dating from the two
centuries immediately preceding the Christian era.[1] Yet, poor as it
was, its art is distinct. There one recognizes all that general delight
in decoration and that genuine artistic instinct which mark Late Celtic
work, while the technical details of the ornament, as, for example, the
returning spiral, reveal their affinity with the same native fashion. On
the other hand, no trace of classical workmanship or design intrudes.
There has not been found anywhere in the village even a _fibula_ with a
hinge instead of a spring, or of an Italian (as opposed to a Late
Celtic) pattern. Turn now to the Romano-British villages excavated by
General Pitt-Rivers at Woodcuts and Rotherley and Woodyates, eleven
miles south-west of Salisbury, near the Roman road from Old Sarum
(Sorbiodunum) to Dorchester in Dorset.[2] Here you may search in vain
for vestiges of the native art or of that delight in artistic ornament
which characterizes it. Everywhere the monotonous Roman culture meets
the eye. To pass from Glastonbury to Woodcuts is like passing from some
old timbered village of Kent or Sussex to the uniform streets of a
modern city suburb. Life at Woodcuts had, no doubt, its barbaric side.
One writer who has discussed its character with a view to the present
problem[3] comments, with evident distaste, on 'dwellings connected with
pits used as storage rooms, refuse sinks, and burial places' and
'corpses crouching in un-Roman positions'. The first feature is not
without its parallels in modern countries and it was doubtless common in
ancient Italy. The second would be more significant if such skeletons
occupied all or even the majority of the graves in these villages.
Neither feature really mars the broad result, that the material life was
Roman. Perhaps the villagers knew little enough of the Roman
civili