4 ‘minims,’ and 3 illegible. One of the coins of Maximin was lying close to the foundations at a considerable depth.
   Some historical conclusions can be drawn from these facts. We may distinguish two periods. In the first period, which perhaps included the second century, the spot was a harbour occupied by part of the classis Britannica. A praefect of this fleet erected an altar, and there were buildings constructed with tiles stamped with the mark of the fleet. This period passed away. The altar was thrown down, and lay a while under water—though this, indeed, might have happened at any time. Still more significant, the tiles bearing the fleet stamp ceased to be made. Then came the second period. The altar and tiles and various stones from buildings of the first period were used to construct a massive fortress-wall with gates and towers in the manner of the fourth century. Coins, too, betoken an occupation which began near the opening of this century. We may perhaps advance further. The remains of the first period are not in situ ; those of the second, coins as much as stones, seem to belong to the spot where they were found. It is therefore possible that the occupation of the first period covered a slightly different site, and that this is the reason why the coins found in the fort include so few of the earlier period.
   With these deductions we may compare the few facts known about the place from ancient books. Its position at the end of Stone Street and its name combine to identify it with the portus Lemanis mentioned in the Itinerary as sixteen miles from Canterbury, and therefore with the misspelt Lemavio of the Peutinger map, and the Lemanis which the Ravennas puts next to Dubris (Dover) and Duroaverno Cantiacorum (Canterbury). Further, we may cite here the ‘Notitia,’ which states that a ‘ numerus Turnacensium’ garrisoned ‘Lemannis’ in the fourth century.⁹⁶  With this further aid we can briefly sketch the history of Roman Lympne. At first, during some part of the first three centuries, it was a harbour, used little for trade, perhaps, but forming, like Dover, a station of the British fleet. Later, towards the year 300, the fleet vanishes. Instead, we find a fort erected probably near, but not on the precise spot occupied by the fleet. It was garrisoned by troops. Yet it came down to the water, and we can hardly suppose that the forces which it accommodated acted only on dry land.^96a  Here, as in the other forts of the Saxon Shore, and perhaps here even more clearly than elsewhere, we see that while the classis Britannica of the first three centuries disappears about A.D. 300, and the system which succeeded was ostensibly a system of land-defence, some use nevertheless was still made of ships.
  ⁹⁶ Itin. .Ant. 473, 7 and 10; Ravennas, 428, 2; Notitia occ. xxvii, 5. Probably the river Lemana of Rav. 438, 19, should be also connected, as most writers think. The nominative of the name is generally taken to have been Lemanae, but it does not occur.
   ^96a For possible roads connecting the forts at Lympne and Pevensey, see below, p: 139, and Lympne with Dover, p. 140. For an altar, possibly from Lympne, see below, p. 169, s.v. Stone-in-Oxney.