Now, the way from the right bank of the Somme to Boulogne is interrupted by several streams, much the most marshy and broad of which is the Authie. The Romans bridged the Authie at _Ad Pontes_ in the course of their great Trunk Road to Britain, and any way which led from the lowest ford over the Somme to Boulogne would have to join that great Trunk Road before or at the bridge if it were to take advantage, as commerce would have to do, of that sole passage of the very difficult and marshy Authie valley which can nowhere be crossed save upon a causeway. I have in a former page remarked upon the importance of Ad Pontes (the modern Ponches), and pointed out that it gives the whole county its name of Ponthieu. We must expect, therefore, any direct commercial way northward from the ford to make directly for Ponches. To strike the great Trunk Road higher up would be to go out of one's way; to strike it lower down would be to strike the Authie Valley at an impassable point. When an ancient way has disappeared, certain indications of its track, especially as that track may be presumed to be direct, survive, and among these are wayside tombs, parish boundaries, and mills or other places which, for the conveyance of heavy merchandise, are placed near such a road if possible. All these three kinds of indications are available in this particular case. The medieval mill which was so important a monopoly of the medieval community was not built in the most natural place for it, on the summit of the hill just above Port, but some thousand yards and more away down the river bank, and over against it is a group of tombs. Moreover, between the two runs the long north-western boundary of the parish or commune of Port which is prolonged in the boundary of the parish of Sailly.[11] We have here, then, a convergence of proof which confirms the vaguer traditional site, for the end of this line upon the river, passing between the tombs and the old mill, strikes the bank within the limits of distance from Port which were set down in the local notes printed in 1840. But there is more. The forming of successive embankments one below the other for the gradual reclamation of land in the Somme estuary was not an easy matter. They had to be strong to withstand a strong tide, and there was no good bottom to be found in the deep mud of the valley floor. It is a significant evidence of this difficulty that the embankments stand so far apart, and that the last has had to take advantage of the long-established work of the railway viaduct. It is therefore a legitimate conjecture that the hard bottom afforded by the old Blanchetaque would be made use of, and as a fact we find the principal embankment between Port and the sea coinciding exactly with the line established by the tombs, the parish boundaries, and the site of the mill.