the structure really stood near the fort, it may have stood inside it. But we ought possibly to accept the account of Harris, and suppose it to be something separate from the fort, perhaps a farm or villa, half a mile away. The occurrence of mosaic flooring, unusual in forts, favours this view, and an apparent parallel will meet us at Richborough. (2) Battely also mentions some ‘ cisterns,’ 10 ft. to 12 ft. deep and broad, lined with oak stakes and planks, and bottomed with puddled clay. In his opinion they were meant to store rain-water; others have called them rubbish-pits; it is now impossible to refute or accept either conjecture. (3) Hasted alleges that tiles, bricks, and traces of tessellation had been found north of the fort as far as the Black Rock. But he gives no details, and the remains may be those recorded above or merely debris washed out of the north face of the fort.²³
   Of smaller objects we know more. Not a few were discovered in the seventeenth century. Battely describes and illustrates a gold chain with blue beads between the links, a bronze enamelled pendant, a fibula, rings, a statuette of Mars and the head of another statuette, keys, two spoons, the ornamented handle of a clasp-knife, and a ‘strigil’ 12 in. long, all bronze; various pieces of pottery, including Samian and a ruder large four-handled jar holding twenty-four quarts, and some tiles.²⁴  Other recorded objects are an inscribed tile, unfortunately undecipherable, an intaglio of Mercury, a bronze ring and two fibulae (one early and one late in date) in Maidstone Museum, and pottery.²⁵  No inscriptions have been found save the illegible tile, and only six potter’s stamps are recorded (1-5 Samian; 6 on a pelvis or mortarium).
   1.   MARSI.M     Battely,plate ix,p. io6.
   2.   PRIM ITIVI   Ibid. p. 105.
   3.   . . . TACI      Ibid. p. 105, Trin. Coll. Library.
   4.   CCF            Ibid. p. 105.
   5.   OF PRIM     Gent. Mag. 1884 (i), 372, possibly a blundering recollection of No. 1 ;
                                   but the stamp is common.
   6.   LVGVDV    Arch. Journ. xlvii, 234: ‘made at Lugudunum, in Gaul’ : i.e. an import,
                                   like the Samian ware.
In general, little of any sort has come to light recently, and Dowker states that no discoveries occurred when some coastguards’ cottages were built within the fort.
   Coins have turned up in abundance. Leland testifies to ‘much Romain mony,’ Duncombe to ‘vast quantities of Roman coins, chiefly of the Lower Empire,’ and Battely to coins of all periods from Cesar to Honorius. The finds that are actually specified seem to fall into two portions. The first comprises a few early issues—five or six British gold, mostly contemporary
    ²³  Hasted, iii, 634.
    ²⁴  Battely, plates ix—xiii, repeated by Harris, p. 24.8, and Gent. Mag. 1774, p. 353; C. R. Smith, Richborough, etc., pl. vii. The library of Trinity College, Cambridge, possesses Battely’s collection. Personal examination has shewn that some of the articles are labelled ‘Richborough’ in error. An inscribed ring (Corpus Inscr. Lat. vii, 24*, C. W. King, Antique Gems, p. 290), some’ scale beams,’ and one or two other things, also preserved at Cambridge, are plainly medieval, though the ‘scale-beams’ are figured by Gough (Adds. to Camden, i, 344, plate xvii) as Roman.
   ²⁵  For the tile see Arch. viii, 79, with plate: hence Corpus Inscr. Lat. vii, 1261.  For the intaglio, C. R. Smith, Richborough, etc., pl. vii, fig. i6, and Brit. Arch. Assoc. Journ. xiv, 95. For other pottery, Soc. Ant. Minutes, 12 June, 1735, i8 Dec. 1735,  Arch. Journ. xlvii. 234, etc. Leland mentions a ‘ christal stone’ engraved CLAVDIA ATE PICCVS, which adorned the binding of an ancient book of the Gospels in the church in his time, and Hübner admits it (Corpus Inscr. Lat. vii, 1325). Probably it is Roman work, and as Atepiccus is a Celtic name, it may belong to Gaul or Britain; but it has no demonstrable connexion with Roman Reculver.