Such a connexion does not in itself imply a pre-Roman occupation of this particular spot. In well-known instances on the Continent—at Gergovia and Bibracte—the Romans compelled or encouraged the Gallic tribesmen to leave their old hill-capitals and dwell in new towns on the plain. With these analogies in mind, we might point to the ramparts of pre-Roman type which surround the hilltop of Bigbury Wood, two miles to the westward.⁵  But before we suggest that the Romans removed the natives of Bigbury down to the Romano-British town beside the Stour, we need to know more of that ‘camp.’ Indeed, no instance of this process of transmigration has yet been adequately demonstrated in Britain. Here and there we have a hint of it; for example, excavation has made it tolerably clear that the pre-Roman chef-lieu of the Silures, west of the Severn, must be sought elsewhere than on the site of their Romanizing capital at Caerwent. On the other hand, at Silchester and Winchester there is some reason for supposing an unbroken continuity between the pre-Roman and Roman tribal townships.⁶  In short, the problem is one which must be solved afresh for each individual site. Its solution can be found only by deep and careful excavating ; and at Canterbury we are confronted at the outset alike with a complete absence of scientific digging and with a generally indifferent record of chance-discoveries (P1. XII).
   Let us now turn to these discoveries. Amongst them evidences of pre-Roman life in Canterbury are scanty. What we have seems to belong mostly to the south side of the town. There a few relics of the Stone and Bronze Ages have come to light, mostly in or near Wincheap. Two fibulae of Italian character and approximately of the 7th century B.C., now preserved in the Canterbury Museum, belong to a considerable class of these objects preserved in our various museums and rarely of authenticated origin. Better evidence is that of a Gaulish and three late British coins,⁷  although, in the absence of other pre-Roman relics, it is well to remember that such coins have on occasion been found with Roman coins as late as those of Domitian.⁸  These meagre discoveries with others of even vaguer character are insufficient in themselves to suggest prehistoric settlement on the site. Equally insignificant in this connexion is the reputed discovery of a socketed axe of late Bronze Age type in one of a group of mounds which lay close within and without the line of the medieval walls in the south-eastern quarter of the city.⁹  These mounds were probably themselves of Roman date and will be discussed below (p. 77).
   We may say, therefore, that tangible evidence for a prehistoric settlement here is lacking. One other consideration, and one only, may be thought to have some slight weight in relation to this problem. A glance at the map will show that all the Roman main roads from Richborough, Dover, Lympne and London change direction at this spot. In the apparent absence of determining geographical factors, this somewhat suggests that when the road-system was laid out, at the beginning of the Roman occupation, Canterbury was already of sufficient prestige to form a natural focus. This possibility may be noted, but, without further evidence, cannot take us far. It is insufficient to materialize a pre-Claudian Canterbury.
   ⁵   See Arch. Cant. ix, 13
   ⁶   For a general discussion of the tribal system in Roman Britain, see Haverfield and Macdonald, The
Roman Occupation of Britain, pp. 190 ff.
   ⁷   J. Evans, Brit. Coins, pp. 203, 478, 482, 527.   ⁸ Numismatic Chronicle, 4 ser. viii, 1908, p. 81.
   ⁹  J. Evans, Bronze Implements, fig. 118; Brent, Cant. in Olden Time, pp. 5, 48, p1. iii.