was an attractive one. It is a small, flat
promontory of clean chalk and gravel sheltered within a fold
of the Medway and rising slowly southwards to the Kentish
hills. With water or marshland on three sides of it, defence
did not present any special difficulty, although from a
military point of view the site is not notably strong. Indeed,
its landward accessibility is, in some degree, rivalled by its
accessibility from the sea, and the modern development of the
district might tempt us to compare its position at the head
of. ‘the Medway estuary with that of the port of. London
near the former limit of the tidal Thames. The Medway estuary,
however, is a dangerous one, and, besides, a harbour at this
point can have, had no great attractions for the men of
Romano-British times. The principal routes which then
connected. ‘south-eastern Britain and the Continent were
either a short sea passage to an East Kent port with a journey
by land, or, in the alternative, a long sea: voyage direct to
London. The use of Rochester as a port. would have meant a
fairly long sea voyage with an appreciable land journey in
addition, and this suited neither quick nor slow, traffic, We
must therefore suppose that it was rather the other advantages
of the site which drew settlers to the spot—the North Kent
roadway, the facilities for a bridge, and perhaps. more the
combination of these facilities than any one of them singly.
The name of this Roman bridge-head settlement. is
not in doubt. The Second Iter of the Antonine Itinerary places
Durobrivae—if that be the nominative .of a. name which
occurs only in the dative or ablative, Durobrivis—at
a distance of 25 miles from Durovernum or
Canterbury, along the main road towards London. This .
distance falls barely a mile short of the map-distance for
Rochester. Again, in the Third and Fourth Itinera, the same
distance is given as 27 (Roman) miles. A garbled version of
the name should perhaps be recognised in the Roribis which,
lies north-west of Durolevo (probably Faversham—p.
93) on the Peutinger Map (P1.VI),
and should, it has been boldly suggested, be restored as [Du]rorbus
;16b whilst the Ravenna Geographer places Durobravis
in his list between Richborough and London. As in the case
of Canterbury (see above, p. 63)—though our evidence is here
less abundant—the Roman name of Rochester seems occasionally
to have survived the introduction of its Saxon. successor,
Rofesceastre; for a Kentish coin of the ninth century bears
the inscription DOROBREBIA CIBIT.
The precise significance of such survivals or revivals is not,
indeed, easy to estimate. They should perhaps be ascribed
rather to literary than to folk tradition, but the question
fortunately does not arise in the present context.
The details of ‘the Romano-British settlement
itself are scantily known. Indeed, but for the observations of
Mr. George Payne and others towards the end of the last
century, we could not attempt to draw any picture of it. The
most striking remains are those of the Roman walls, preserved
to some small extent in the medieval fortifications. The walls
are described variously as 6 ft., 7 ft. and 8 ft. in
,thickness, a variation which may be explained partially by
the presence in at least one stretch17 of an
internal offset, 8 in. wide, at a height of 1 ft. 8 in. above
the base. The best attested measurement gives the width as 6
ft.. 10 in.18 The core, which stands to a
maximum height of 14 ft., is of
16b See O. G. S.
Crawford, Journal of Roman Studies, xiv, 138.
17 Arch. Cant. xxxvii,
lxvi, referring to a discovery along the line of the
18 Ibid. xxix, lxxxiv.