Luckily, we have some indication upon the farther bank.  A way can then be traced here as a lane (and in the gaps as a right of way, as a path, or sometimes only by its general direction) for some miles on the Oxfordshire side as it approaches Goring and the river coming from the Chilterns.  And we know the point at which it strikes the village.  This point is at the Sloane Hotel close to the railway; the inn is actually built upon the old road.  Beyond the railway the track is continued in the lane which leads on past the schoolhouse to the old ferry, where there was presumably in Roman times a ford.  If we accept this track we can conjecture that the vicarage of Streatley, upon the Berkshire bank, stands upon the continuation of the Way, and give the place where the pre-historic road crossed the river with tolerable certitude, though it is, I believe, impossible to recover the half-mile or so which lies between Streatley vicarage and the point where the Wantage Road and the Icknield Way separated upon the hillside above.

If the ford lay here the site was certainly well chosen, just below a group of islands which broadened the stream and made it at once shallower and less swift, acting somewhat as a natural weir above the crossing.

The third crossing place of a great pre-historic road, that of the Watling Street, is believed to correspond with the line of that very ugly suspension bridge which runs from Lambeth to the Horseferry Road in Westminster.  This is, according to the most probable conjecture, the place at which the great road which ran from the Straits of Dover to the north-western ports of the island crossed the Thames.

Here, of course, there could be no question of a ford; there can only have been a ferry.  Such a ferry existed throughout the Middle Ages and up to the building of Westminster Bridge, and produced a large revenue for the Archbishop of Canterbury.  The memory of it is preserved in the name of the street upon the Middlesex shore.  The Watling Street is fairly fixed in all its journey from the coast to the Archbishop’s palace on the banks of the river.  On the Middlesex shore it is lost, but it may be conjectured to have run in a curve somewhere in the neighbourhood of Buckingham Palace up on the higher ground west of the Tybourne, parallel with or perhaps identical with Park Lane until we find it certainly again at the Marble Arch, whence in the form of the Edgware Road it begins a clear track across North-Western England.

As for the Fosse Way, it only just touches the valley of the Thames.  It crosses the line of the river in a high embankment a mile or so below its traditional source at Thames head, but above the point where the first water is seen.  A small culvert running under that embankment takes the flood water in winter down the hollow, but no longer covers a regular stream.

Besides these four crossings of the old British ways above London Bridge there is the crossing of the Roman Road at Staines, which may or may not represent a passage older than the Roman occupation.  We have no proof of its being older.  The river is deep, and, unless the broken causeway on the Surrey shore is regarded as the remains of British work, there is no trace of a pre-Roman track in the neighbourhood.