temples were of the well-known Romano-Celtic type, with square cella and surrounding portico. No indications as to their date or dedications were found during the hurried examination of them.
   Beyond the temples, during the construction of the main railway line in 1846 at the south of the cliff, 600 yards from the fortress and only a little above marsh level, a portion of another building was similarly destroyed. No more was uncovered than some flint-and-tile walling and parts of rooms (one apsidal) paved with concrete—all suggestive of a dwelling-house or baths.³⁸  It may remind us of a similarly placed building at Reculver (p. 21). Some Samian ware, now in the Liverpool Museum, was found at the same time.
   Lastly, one structure has been found actually on marsh level. This came to light in the eighteenth century, about 200 yards north of the fort and at the foot of the cliff. It was 4 ft. high, triangular in shape, measuring 10 ft. each way and pointing an angle towards the sea, and consisted of a brickwork shell, two bricks thick, filled with earth.³⁹  This can scarcely have been a tomb. Boys called it a wharf. It may rather remind us of the solid hexagonal Lantern of Augustus’ at the entry to the port of Fréjus in Provence, but that structure is equally a mystery and, therefore, does not help.
   Of the cemeteries of the place almost nothing is known. Outside the fortress a few urns with burnt bones have been found near the farmhouses to the west. Within the fortress many human skeletons, bones and ashes have been found in the eastern part close to the cliff. Here in the Middle Ages stood a chapel dedicated to St. Augustine,⁴⁰  and most of the skeletons are doubtless connected with this. But the ashes must be earlier and may well antedate the fortress.⁴¹  One tomb at least stood upon the site before the days of the Saxon Shore. In 1927, under the west wall of the fortress, just south of the gateway, was found a skeleton which had been buried in a coffin. There was no grave-furniture except a bronze pin, which had probably fastened the shroud. About 6 ft. above the body, at what must have been the ground-level at the period, were the remains of a small masonry chamber, which had been partly cut away when the wall of the (later) fortress was built. There were slight indications that the chamber had been covered by a tumulus, as in a somewhat similar example at Rougham in Suffolk. The date of the burial has been proved to be later than about 200 A.D., as indeed we should expect it to be from its character. It is certainly earlier than the latter part of the third century, when the fortress was built. The first half of the third century is therefore the most likely period.
   If we know little of the cemeteries of Richborough we do not know much more of the road or roads which gave access to the spot. That a road existed we may feel sure. Richborough was not only a fort. It was also a harbour where men and goods were landed and embarked. That does not suit an island divided by water from the roads of Kent. Had Richborough been inaccessible by land, the landing-place would have been on the Kentish shore of its harbour and not on its island. Of such a landing-place, with its houses and inhabitants, there is no trace anywhere save on Richborough hill, and
   ³⁸  C. R. Smith, Richborough, etc., 54. (plan).
   ³⁹  Boys, Sandwich, 868; hence C. R. Smith, Richborough, etc., 53.
   ⁴⁰ The very fragmentary remains of this chapel have been excavated and are described in the Second
             Richborough Rep. (Soc. Antiq.), p. 34.
   ⁴¹ T. C. Bell, Arch. Aeliana, ii (1852), 377.