A common method among those who belittle the continuity of our civilization, is to deny a Roman origin to any town in which Roman remains do not happen to have been noted as yet by antiquarians. Even under that test we can be certain that Windsor, Lewes, Arundel, Dorking, and twenty others, were seats of Roman habitation, though the remaining records of the first four centuries tell us nothing of them. But in nine cases out of ten the mere absence of catalogued Roman remains proves nothing. The soil of towns is shifted and reshifted continually generation after generation. The antiquary is not stationed at every digging of a foundation, or sinking of a well, or laying of a drain, or paving of a street. His methods are of recent establishment. We have lost centuries of research, and, even with all our modern interest in such matters, the antiquary is not informed once in a hundred times of chance discoveries, unless perhaps they be of coins. When, moreover, we consider that for fifteen hundred years this turning and returning of the soil has been going on within the municipalities, it is ridiculous to affirm that such a place as Oxford, for instance—a town of importance in the later Dark Ages—had no Roman root, simply because the modern antiquary is not yet possessed of any Roman remains recently discovered in it: there may have been no town here before the fifth century: but it is unlikely. One further point must be noticed before we leave this prime matter: had there been any considerable destruction of the Roman towns in Britain, large and small, we should expect it where the pirate raids fell earliest and most fiercely. We should expect to find the towns near the east and the south coast to have disappeared. The historical truth is quite opposite. The garrison of Anderida indeed and of Anderida alone (Pevensey) was, if we may trust a vague phrase written four hundred years later, massacred in war. But Lincoln, York, Newcastle, Colchester, London, Dover, Canterbury, Rochester, Chichester, Portchester, Winchester, the very principal examples of survival, are all of them either right on the eastern and southern coast or within a day’s striking distance of it. As to decay, the great garrison centre of the Second Legion, in the heart of the country which the pirate raiders never reached, has sunk to be little Caerleon-upon-Usk, just as surely as Dorchester on the Thames, far away from the eastern coast, has decayed from a town to a village, and just as surely as Richboro’, an island right on the pirate coast itself, has similarly decayed! As with destruction, so with decay, there is no increasing proportion as we go from the west eastward towards the Pirate settlements. But the point need not be labored. The supposition that the Roman towns disappeared is no longer tenable, and the wonder is how so astonishing an assertion should have lived even for a generation. The Roman towns survived, and, with them, Britain, though maimed.