Supposing, then, the fleet of Cæsar at anchor at a distance of half a mile opposite Dover, as it experienced the effect of the shifting of the currents towards half-past three o’clock in the afternoon, the question becomes reduced to that of determining the day of the end of the month of August when this phenomenon took place at the above hour. We know that in the Channel the sea produces, in rising and falling, two alternate currents, one directed from the west to the east, called _flux_ (_flot_), or current of the rising tide; the other directed from the east to the west, named _reflux_ (_jusant_), or current of the falling tide. In the sea opposite Dover, at a distance of half a mile from the coast, the flux begins usually to be sensible two hours before high tide at Dover, and the reflux four hours after. So that, if we find a day before the full moon of the 31st of August, 699, on which it was high tide at Dover, either at half-past five in the afternoon or at midday, that will be the day of landing; and further, we shall know whether the current carried Cæsar towards the east or towards the west. Now, we may admit, according to astronomical data, that the tides of the days which preceded the full moon of the 31st of August, 699, were sensibly the same as those of the days which preceded the full moon of the 4th of September, 1857; and, as it was the sixth day before the full moon of the 4th of September, 1857, that it was high tide at Dover towards half-past five in the afternoon (_see the Annuaire des Marées des Côtes de France for the year_ 1857),[384] we are led to conclude that the same phenomenon was produced also at Dover on the sixth day before the 31st of August, 699; and that it was on the 25th of August that Cæsar arrived in Britain, his fleet being carried forward by the current of the rising tide. This last conclusion, by obliging us to seek the point of landing to the north of Dover, constitutes the strongest theoretic presumption in favour of Deal. Let us now examine if Deal satisfies the requirements of the Latin text. The cliffs which border the coasts of England towards the southern part of the county of Kent form, from Folkestone to the castle of Walmer, a vast quarter of a circle, convex towards the sea, abrupt on nearly all points; they present several bays or creeks, as at Folkestone, at Dover, at St. Margaret’s, and at Oldstairs, and, diminishing by degrees in elevation, terminate at the castle of Walmer. From this point, proceeding towards the north, the coast is flat, and favourable for landing on an extent of several leagues.