The prehistoric road as we know it went then at last in a great flat curve from Winchester to Canterbury, following the simplest opportunities nature afforded. It went eastward, first up the vale of the Itchen to the watershed, then down the vale of the Wey, and shortly after Farnham struck the range of the North Downs, to which it continued to cling as far as the valley of the Stour, where a short addition led it on to Canterbury. Its general direction, therefore, when it had settled down into its final form, was something of this sort:-- [Illustration] When Winchester began to affirm itself as the necessary centre of south England--that is of open, rich, populated, and cultivated England--the new tributary road would rapidly grow in importance; and finally, the main traffic from the western hills and from much of the sea also, from Spain, from Brittany, and from western Normandy, probably from all southern Ireland, from the Mendips, the south of Wales, and the Cornish peninsula, would be canalised through Winchester. The road from Winchester to Farnham and so to Canterbury would take an increasing traffic, would become the main artery between the west and the Straits of Dover, and would leave the most permanent memorials of its service. * * * * * Winchester and Canterbury being thus each formed by the sea, and each by similar conditions in the action of that sea, the parallel between them can be drawn to a considerable length, and will prove of the greatest value when we come to examine the attitude of the Old Road towards the two cities which it connects. The feature that puzzles us in the approach to Canterbury may be explained by a reference to Winchester. An unsolved problem at the Winchester end may be referred for its solution to Canterbury, and the evidence of the two combined will be sufficient to convince us that the characters they possess in common are due to much more than accident. Of all the sites which might have achieved some special position after the official machinery of Rome with its arbitrary power of choice had disappeared, these two rose pre-eminent at the very entry to the Dark Ages, and retained that dual pre-eminence until the great transition into the light, the Renaissance of civilisation at the end of the twelfth century and the beginning of the thirteenth. For six or seven hundred years the two towns were the peculiar centres of English life. Winchester was a capital longer than London has been; Canterbury ruled the religion of this island for over nine hundred and forty years. So much we know for certain, and more may be presumed; but I have conjectured that these sites were of equal importance before the advent of Rome, and such a conjecture needs support. I will maintain that the barbaric centuries which followed the decline of the Empire reproduced in Britain original conditions, and restored their value to sites neglected during the period of Roman order.