_First._--The spur, which leaves the main range of hills for the river (such a spur as has elsewhere, at Shalford, and at Dorking, and at Otford, attracted the Old Road towards the ford it points to), touches indeed both Snodland and Lower Halling on either side, but with this great difference--that Snodland is on the south, Lower Halling upon the north of the ridge. The elevation is not pronounced, the slope is slight, but a little experience of such ground at various seasons will determine one that the southern bank would be chosen under primitive conditions. In such a conformation the southern bank alone has during the winter any chance of drying, and in a dry summer, it matters little whether a slope be partly of clay[35] (as is the descent to Snodland) or of chalk (as is that to Lower Halling). During more than half the year, therefore, the descent to Snodland was preferable; during the other half indifferent. _Secondly._--Immediately before and beyond the Lower Halling crossing no antiquities of moment have been discovered: a grave, possibly Roman, is, I believe, the only one. At Snodland, and beyond its crossing, they are numerous. An ancient and ruined chapel marks the descent from the hills. The church itself has Roman tiles. Beyond the river, the Roman villa which was unearthed in 1896 by Mr. Patrick is precisely upon the road that would lead from such a crossing up to the Pilgrim's Way upon the hill. Close by the origin of this lane from the ford to the hillside were discovered the fragments of what some have believed to be a Mithraic temple; and earlier, in 1848, Roman urns and foundations were found near the road at Little Culand. _Thirdly._--The crossing at Snodland is shallower than that at Lower Halling, and (though I do not pretend that the artifice is prehistoric) the bottom has been artificially hardened. _Fourthly._--There stands at Snodland a church, past the _southern_ porch of which goes the road, and when the river is crossed, and the same alignment followed along the bank upon the further side for a little way, the track again passes by a church, and again by its _southern_ porch. _Fifthly._--The 'Horseshoe Reach'--the reach, that is, between Snodland and Burham--has always marked the limit between Rochester's jurisdiction over the lower, and Maidstone's over the upper, Medway. This is of great importance. All our tidal rivers have a sea-town and a land-town; the limits up to which the seaport has control is nearly always the _traditional crossing-place_ of the river. Thus Yarmouth Stone on the Yare divides the jurisdiction of Norwich from that of Yarmouth; it is close to the Reedham Ferry, which has always been the first passage over the river. For London and the Thames we have the best example of all--Staines.