A HISTORY OF NORFOLK 7 by 1 1 feet in extent, walled with flanged tiles for part of its height, and containing bones of oxen, pigs and other animals, oyster-shells and potsherds of Roman date. Skeletons, coins (mostly a.d. 270-320), tiles and pottery were noticed at the same time in other parts of the field, and a subsequent writer states vaguely that foundations of buildings exist here. (2) In 1843 a claypit north-east of the church yielded a burial urn full of bones and ashes, covered by a tile, and also remains of an inhumation in a wooden coffin. (3) In 1851 a kiln containing broken pottery and a curious iron drying-stand was dug out of a sandpit on the south side of the church, between it and the marshes. (4) In 1855, during the construction of the reservoir, much broken pottery (including Samian), a little bronze bust (? a Faun), a bronze pin, many 'Third Brass ' coins, bones and oyster-shells were found — apparently the filling of a rubbish-pit. (5) In 1879 a rude flint pavement or foundation was exposed in the rectory garden, and some coins, pottery and an ashpit dis- covered. Besides these definitely recorded finds, coins, potsherds and occasionally other objects — such as a little bronze figure of a wolf or dog, formerly the top of a staff or the like — have been recorded without detail of place or circumstance, ever since the days of Sir Thomas Browne. The earliest coins date from about a.d. 80, but those of A.D. 250—380 are the commonest.' These remains, like the remains of Caister-by-Norwich, do not indicate military occupation, and in the entire absence of military in- dications, it will be safer to hold Caister-by-Yarmouth to have been a small settlement of some sort, not adequately known as yet. Its situation may seem exposed to any chance sea rover, but girt as it is with marsh on south and east, it is less exposed than might at first sight be thought, and not more exposed than other Roman remains of a non-military kind on some of our coasts. Excavation or chance discovery can alone, however, decide beyond controversy : it is enough here to state the probabilities. 4. The Villas We have seen that Roman Norfolk is poor in towns. It is hardly richer in villas. No county in the south-east of England possesses, in proportion to its area, so few remains that indicate the residences of land- owners or the scattered structures connected with their estates. Nor are the remains merely few in number : they are also plain in character, and the list which I am about to give contains not even one instance of that common luxury, a mosaic pavement. This poverty is characteristic. We must indeed allow something in our estimate for evidence that has perished and evidence that has not been yet unearthed. The rural
- For (i) see Gentleman's Magazine, 1837, ii. 518 ; for (2) Archaok^cal Journal, iii. 251 and
Dawson Turner MS. 23,027, p. 132 ; for (3) 'Norfolk Archaologj, iv. 352, Journal of the British Archae- ological Association, xxxvi. 90, 206 ; for (4) Norfolk Archaeology, vii. 12 ; for (5) ibid. ix. 361 and Journal of the B. A. A., xxxvi. 89 ; for the bronze wolf (?) Norfolk Archaeology, vii. 356. It and other objects from the site are in the Fitch room of the Norwich Museum. See also Browne's Umhurial, chaps, ii., iii., and the Winchester volume of the Archsological Institute, p. xl. (Samian and other pottery). 294
