^{page 277 note 1} See pp. 277–80 below.

^{page 277 ntoe 2} J.B.A.A. xvi (1953), 6–9.Google Scholar For another modern discussion see Boon, George C., Roman Silchester (London, 1957), pp. 128–3; and edition 2 (1974), pp. 173–83.Google Scholar

^{page 277 note 3} J.R.S. lii (1962), 185–6.Google Scholar

^{page 277 note 4} Archaeologia, liii (1893), 563–8. Plan of building at fig. 4 on p. 564, plan of Insula IV at pl. XLI, and colour reproduction of the mosaic panel by George E. Fox at pl. XL.Google Scholar

^{page 278 note 1} [Richmond's plan (fig. 2) gives an overall measurement of 42 ft. 6 in. (12․96 m.). See p. 280 n. 1 for statement of overall dimensions.]

^{page 278 note 2} [The 1961 excavations showed that this is not the case.]

^{page 280 note 1} The building is 42 ft. 6 in. (12․96 m.) long overall, or 36 ft. 6 in. (11–12 m.) without the apse. The width of nave and aisles overall is 28 ft. (8․54 m.), and that of nave and transepts 31 ft. (9․45 m.).

^{page 285 note 1} See Richmond's remarks in his chapter ‘Roman Timber Buildingi in Jope, E.M.Studies in Building History (London, 1961), p. 23Google Scholar; cf. also Arch. Ael⁴ xxxi (1953), 210 and fig 7.Google Scholar

^{page 288 note 1} J.R.S. lii (1962), 186.Google Scholar

^{page 288 note 2} As in the St. George's Street, Canterbury, bathhouse; cf. J.R.S. xxxviii (1948), 98Google Scholar; Frere, S. S., Roman Canterbury (3rd edition) (Canterbury, 1962), figs. 15, 16.Google Scholar

^{page 289 note 1} I am grateful to Dr. Richard Reece for advice on this, He writes: ‘it is very difficult to describe all seven of these coins, which stretch between 270 and 360, as a single group. The main problem, apart from the absence of comparable groups, is one of size, for the three coins of 309–24 are substantially larger, heavier, and probably had a much better silver surface, than the others. It is very rare to find such a proportion of these coins in deposits of c. 340 (the date to which this group could be assigned if the latest coin were intrusive), and the same is true of deposits of c. 360 (the date if the latest coin is integral). Groups of coins from burials in the Lankhills cemetery, Winchester, show that surprising combinations of coins of different dates are possible in the later fourth century; but in these odd groups, though the dates of the coins are disparate, the sizes are uniform. It would therefore require special pleading to describe this collection as a hoard, and still more as the contents of a purse, as the earliest Constantinian coin is unworn and the latest coin of all is worn. From the numismatic point of view, it remains a problem.’

^{page 290 note 1} J.R.S. lii (1962), 186.Google Scholar

^{page 290 note 2} The flints resemble the heavy bedding sometimes found beneath an opus signinum floor, e.g. in a bath-building: but there was no trace of any applied surfacing of such a sort.

^{page 291 note 1} The permanent water-table was deeper than this. Wells at Silchester vary between c. 8 and 30 ft. (2․5–9 m.) deep: Boon, op. cit. (1957), p. 159.

^{page 291 note 2} Archaeologia, liii (1893), 567 quoted above, p. 280. Boon, op. cit. (1957), pp. 135 f., suggests that this was an early timber house.Google Scholar

^{page 291 note 3} Archaeologia, liii (1893), 564 (cited p. 278 above).Google Scholar

^{page 291 note 4} And a sherd of Trier samian ware of the third century in the filling of a post-hole, if this can be attributed to the same building (p. 299); note also the coins of Victorinus already mentioned in the well—with the same proviso.

^{page 292 note 1} M., R. E. and Wheeler, T. V., Verulamium, A Belgic and Two Roman Cities (Oxford, 1936), pp. 108Google Scholar, 148, pl. XLVIII B. The floor at Wroxeter mentioned above, p. 280, is not dated but may be assumed also to be of the fourth century. Its pattern is similar to that at Veru-lamium but designed with greater regularity, and it is much larger (c. 8 ft. x 11 ft. = 2․44 x 3․35 m.). For illustration see J.B.A.A. xvii (1861), pl. 9. A somewhat analogous pattern is known from the late first-century palace at Fishbourne; Cunliffe, Excavations at Fishbourne 1961–1969 (London, 1971), pl. LXXIV c and p. 147.Google Scholar

^{page 292 note 2} The main points were (a) the condition of the floor with its suggestion that the priest would have celebrated mass facing west; (b) lack of trace of, or space for, a synthronos in the apse; (c) absence of any Christian object; (d) resemblances to known pagan shrines, e.g. the underground basilica near the Porta Maggiore, Rome.

^{page 292 note 3} Fremersdorf, F., Altestes Christentum (Berlin, 1956)Google Scholar, Taf. 17–20; Krämer, W. (ed.), Neue Ausgrabungen in Deutschland (Berlin, 1958), pp. 329–39Google Scholar; Toynbee, J. M. C., J.B.A.A.³ xvi (1953), pl. 1. The original simple rectangular chapel with western apse was built c. A.D. 320; towards the end of the century it was enlarged by two aisles and a narthex. The first chapel, measuring 36 x 30 ft. without the apse, is almost the same size as the Silchester building at 36․5 x 28 ft.Google Scholar

^{page 293 note 1} Codex Iustinianus, 1, viii: … signum salvatoris Christi nemini licere vel in solo vel in silice vel in marmori-bus humi positis insculpere vel pingere, sed quodcumque reperitur tolli:…

^{page 293 note 2} The foundation for a statue-base is a less likely explanation, for this would be more likely to be placed against the wall.

^{page 293 note 3} Testamentum Domini nostri Iesu Christi, ed. Ignatius Ephraem II (Rahmani), patriarcha Antiochenus (Mainz, 1899), § 19. Translation by Cooper, J. and Maclean, A. J.The Testament of our Lord (Edinburgh, 1902).Google Scholar For a discussion see Morin, D. G., Revue Bénédictine, xvii (1900), 10–28.CrossRefGoogle Scholar

^{page 294 note 1} See Krautheimer, R., Early Christian and Byzantine Architecture (Harmondsworth, 1965), p. 211.Google Scholar

^{page 294 note 2} Med. Arch, xv (1971), 4.Google Scholar

^{page 294 note 3} Noll, R,FrühesChristentumin = sterreich (Vienna, 1954).Google Scholar

^{page 295 note 1} For some western provincial examples see Brown, P. D. C., Britannia, ii (1971), 225–31CrossRefGoogle Scholar; others more generally in Actas del VIII Congresso International de Arqueologia Cristiana (Rome and Barcelona, 1972), passim; de Palol, P.Arqueologia Cristiana de la Españna Romana (Madrid, 1967), pp. 147–82Google Scholar; de Palol, P. (ed.), I^a Reunion Nacional de Arqueologia Paleocristiana (Vitoria, 1967), passim.Google Scholar

^{page 295 note 2} When this difficulty was raised in discussion after his lecture to the Society, Richmond said he envisaged some form of light or temporary wooden screen. But even this would be unparalleled.

^{page 295 note 3} Med. Arch, xv (1971), 3.Google Scholar

^{page 295 note 4} Goodenough, E. R., Jewish Symbols in the Greco-Roman Period (New York, 1953), i, 241: iii, fig. 631; E. L. Sukenik, The Synagogue of Beit-Alpha (1932); another plan in Zev Vilnay, Guide to Israel (9th ed., 1966), p. 394.Google Scholar

^{page 296 note 1} Applebaum, S., ‘Were there Jews in Roman Britain?', Trans. Jewish Historical Society of England, xvii (1953), 189–205.Google Scholar

^{page 296 note 2} Eusebius, Ecclesiastical History, book x.

^{page 296 note 3} Ibid, x, 39–40.

^{page 296 note 4} Information from Dr. A. Negev, the excavator.

^{page 296 note 5} Crowfoot, J. W., Churches at Jerash(British School at Jerusalem, 1931), p. 8Google Scholar; Kraeling, C. H., Gerasa, city of the Decapolis (New Haven, Connecticut, 1938), pp. 210 ff.Google Scholar

^{page 296 note 6} Forschungen in Ephesos (Vienna, 1932), iv (i), 42Google Scholar; plan in Krautheimer, R., Early Christian and Byzantine Architecture (Harmondsworth, 1965), p. 81, fig. 27.Google Scholar

^{page 296 note 7} Gsell, S., Monuments antiques de I'Algérie (Paris, 1901), p. 268Google Scholar; cf. de Roch, S., Tébessa, antique Theveste (Algiers, 1952), plan.Google Scholar

^{page 296 note 8} Paulinus, Epist. xxxii, 15; cf. Natal, s. Felicis: Patrologia Latina, Ixi, col. 671 f.

^{page 296 note 9} Paulinus, Epist. xiii, 13:… ubi cantharum ministra manibus et oribus nostris fluenta ructantem fastigiatus solido aere tholus ornat et inumbrat, non sine mystica specie quattuor columnis salientes aquas ambiens.

^{page 296 note 10} Krautheimer, op. cit., pl. 7. Cf. also Cabrol, F., Dictionnaire d' Archéologie chrétienne, ii, col. 1958 if.Google Scholar

^{page 296 note 11} De Rossi, Inscript. Christ, urbis Romae ii (i), p. 80, no. 13.