Again, that all-important consideration in a new country--I mean the dryness of the soil over which a road passes--was very much helped by taking the more direct of the two lines. Ground with some slope to it, and always fairly dry, comes here on either side, close to the river. But down by Burford, on the western side, there is quite a little plain, which must have been marshy, and for all I know, may be so still. Moreover, we shall find further on at Otford, and at Snodland, that the Old Road in crossing a valley always chooses a place where some spur of high land leads down to the river and corresponds to a dry rise immediately upon the other bank. Coupled with the fact that a direction such as I suggest makes a natural link between the two known parts of the road, being very nearly in one alignment with them, and remembering that Burford Bridge was built in connection with a very much later Roman way northward up the valley (it is evidently the Bridge of the Stane Street), that its direction and the place at which it crosses are obviously dependent on a north and south road, not on an east and west one, we decided to approach the eastern bank, and to see whether any existing trace of a road would support what seemed to us the most tenable hypothesis. It cannot be pretended that any very distinct evidence clinched what remains, after all, our mere theory; but there was enough to convince us, at least, and I believe to convince most people, who will do as we did, and stand upon the Old Road at the base of Box Hill looking towards Denbies on the other side of the valley. He will see, following a very obvious course, a certain number of yews of great age, remaining isolated in the new-ploughed land. These, leading across the river, are continued at a very slight angle by a definite alignment of three trees, equally isolated though far less old, standing equally in comparatively modern cultivated land, and leading directly to the place where the track is lost at its exit from Denbies Park; and the whole line follows two spurs of land which approach the river from either side. A conclusion thus reached cannot pretend to such a value as I would demand for the rest of our reconstructions: such as, for instance, can legitimately be demanded for the way in which we filled the gap at Puttenham. But it is far more convincing on the spot, and with the evidence before one (such as it is), than any verbal description can make it, and I would repeat that any one making the experiment with his own eyes will be inclined to agree with us. When we had arrived at this decision in the first hour of daylight we turned eastward, and pursued our way by the raised and yew-lined track which was now quite unmistakable, and which we could follow for a considerable time without hesitation. It ran straight along the 300-feet contour-line, and took the southern edge of a wood called Brockham Warren.