¹ One from Oxborough (TF 7302), reported by Frere, S. S., ‘Roman Britain in 1985: sites explored’, Britannia, xvii (1986), 364–427Google Scholar and pl. XXIX, and, very recendy, another from Reading (SU 7374). For additional information on the Oxborough tank ahead of publication, I thank Mr Christopher Guy of the Trust for Lincolnshire Archaeology. I am indebted to Professor Keith Branigan of Sheffield University for informing me of the find from Reading and to Messrs David Miles and John Moore of Oxford Archaeological Unit for supplying details ahead of publication. This paper forms part of a doctoral thesis for the Department of Classics and Ancient History, University of Queensland, Australia.

² Guy, C. J., ‘Roman circular lead tanks in Britain’, Britannia, xii (1981), 271–6.CrossRefGoogle Scholar

³ Thomas, C., Christianity in Roman Britain to A.D. 500 (London, 1981), 220–7.Google Scholar

⁴ Page, W. (ed.), The Victoria History of the Counties of England: Suffolk, 1 (London, 1911), 309.Google Scholar

⁵ Frere, , op. cit. (note 1), 403.Google Scholar

⁶ Kraay, C. M., ‘An early Christian object from Icklingham, Suffolk’, Antiq. J. xxii (1942), 219–20, andCrossRefGoogle ScholarWest, S., ‘The Romano-British site at Icklingham’, East Anglian Arch, iii (1976), 63–126Google Scholar.

⁷ Guy, C. J., ‘The lead tank from Ashton’, Durobrivae, v (1977), 10–11.Google Scholar

⁸ Curwen, E. C., ‘Roman lead cistern from Pulborough, Sussex’, Antiq. J. xxiii (1943), 155–7CrossRefGoogle Scholar and pl. XXVIII.

⁹ Petch, D. F., ‘A Roman lead tank, Walesby’, Lincolnshire Architect. & Arch. Soc. Rep. & Pap. ix (1961), 13–15.Google Scholar

¹⁰ Not illustrated. This tank, decorated with a lozenge pattern and a single chi-rho, was found in a fourth-century context (D. Miles, pers. comm.). Since, at the time of writing, no photograph was available, only the chi-rho will be discussed in this paper.

¹¹ Guy, , op. cit., (note 2), 273–4.Google Scholar

¹² The chi-rho monogram was probably the most widely used Christian symbol in the fourth century, until it was replaced in popularity by the mono-grammatic cross and the cross standing alone. This development could be attributed to the abolition by Constantine of crucifixion as a death penalty and to the tradition of the discovery of the True Cross by Helena, mother of Constantine (e.g. Nola, Paulinus of, Epistulae, 31.5–6).Google Scholar The use of Christian monograms and crosses has been covered in a well-documented work by Sulzberger, M., ‘La symbol de la croix et les monogrammes de Jésu chez les premiers Chrétiens’, Byzantion, ii (1925), 337–448; see alsoGoogle ScholarCabrol, F. and Leclerq, H., Dictionnaire d'Archéologie Chrétienne et de Liturgie, 15 vols. (Paris, 1920-1953),Google Scholar s.v. ‘croix et crucifix’ and ‘chrisme’, III.

¹³ Thomas, , op. cit. (note 3), figs. 4–7Google Scholar gives a good coverage of these Christian symbols in Britain.

¹⁴ Guy, , op cit. (note 2), 271.Google Scholar

¹⁵ Thomas, , op. cit. (note 3), 221.Google Scholar

¹⁶ Donovan, H. E., ‘Excavations at Bourton on the Water’, Trans. Bristol & Gloucestershire Arch. Soc. lv (1933), 378.Google Scholar

¹⁷ e.g. Rossi, G. B. De, La Roma Sotterranea Cristiana, 3 vols. (Rome, 1864–1877), II, pl. XXXII-XXXIV. 4.Google Scholar

¹⁸ See Justin Martyr, 1 Apology, 60 for a possible early reference. It has been claimed to be the oldest known form of the Christian cross (after the tau cross borrowed from the Old Testament) (Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘croix et crucifix’, III (2), col. 3048); however, Sulzberger, (op. cit., (note 12), 367)Google Scholar believes the two Xs on an inscription from Palmyra, and dating to A.D. 134, are merely punctuation signs. The Christian X is certainly found on early monuments at Chaqqa, in Syria, dating from the fourth century (Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘Chaqqa’, III (1), col. 515). The tau in the Hebrew alphabet was represented by + or X, so there would be no difficulty in believers' equating it with the cross of Christianity.

¹⁹ e.g. Isidore, Origen, 1.3: ‘X quae in figura crucem, et in numero decent demonstrat’.

²⁰ Unpublished, British Museum acquisition no. PRB.1927.1-6.1.

²¹ The Roman Inscriptions of Britain, no. 856.

²² Goodburn, R., The Roman Villa at Chedworth (London, 1979), pl. 12.8.Google Scholar

²³ Thomas, , op. cit. (note 3), 221.Google Scholar

²⁴ Phillips, C. W. (ed.), The Fenland in Roman Times (London, 1970), 208.Google Scholar I am grateful to Dr David Phillipson, University Museum of Archaeology and Anthropology, Cambridge, for photographs of the tanks from Willingham and Wilbraham.

²⁵ Donovan, , op. cit. (note 16).Google Scholar

²⁶ Donovan, H. E., ‘Excavation of a Romano-British building at Bourton on the Water’, Trans. Bristol & Gloucestershire Arch. Soc. lvi (1934), 116–17.Google Scholar

²⁷ My thanks to Mr Brian Dix, Northamptonshire County Council Archaeology Unit, for providing a photograph of this fragment.

²⁸ Webber, F. R., Church Symbolism (2nd rev. edn., Detroit, 1971), 110.Google Scholar

²⁹ Child, H. and Colles, C., Christian Symbols (New York, 1971), fig. 9.Google Scholar It has been pointed out that ‘most of the types of crosses found in Ireland are universal to all Christendom at that time, and have been brought to Ireland from Gaul, Italy or the Near East’: Henry, F., Irish Art in the Early Christian Period (to 800 A.D.) (London, 1965), 54.Google Scholar This would also appear to be the case for Roman Britain.

³⁰ Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘croix et crucifix’, III (2), cols. 3097-102; see also Child, and Colles, , op. cit. (note 29),Google Scholar figs. 15 and 16.

³¹ Ibid., figs, II and 13.

³² While it might be argued that the motif could as easily represent a pagan sun-wheel, it is important to note that each of the four symbols on the tank has been carefully placed so that it forms a cross pattée; there are no vertical or horizontal strokes. This gives further support to a Christian interpretation for the symbol. (Information on the tank kindly supplied by Mrs Elizabeth James, Curator of Lynn Museum, King's Lynn.)

³³ Guy, C. J., ‘A Roman lead tank from Burwell, Cambridgeshire’, Proc. Cambridge Antiq. Soc. lxviii (1978), 2–4.Google Scholar

³⁴ Richmond, I. A., ‘A Roman vat of lead from Ireby, Cumberland’, Trans. Cumberland & Westmorland Antiq. & Arch. Soc. new ser. xlv (1945), 163–71.Google Scholar

³⁵ Donovan, , op. cit. (note 26), 116–17.Google Scholar

³⁶ Ferguson, G., Signs and Symbols in Christian Art (New York, 1961), 153.Google Scholar

³⁷ Webber, , op. cit. (note 28), 363.Google Scholar

³⁸ Child, and Colles, , op. cit. (note 29), 27Google Scholar; this concept was not new: see Isaiah 40:22.

³⁹ Cabrol, and Leclerq, , op. cit. (note 12), s.v. ‘chrisme’, III (1), col. 1501.Google Scholar

⁴⁰ Child, and Colles, , op. cit. (note 29), 27.Google Scholar A similar intensification may be the interpretation of an inscription from Lyons, where the alpha and omega flanking a chi-rho are enclosed in equilateral triangles, the symbol of the Trinity (Corpus Inscriptionum Latinarum, XIII, 2418).

⁴¹ Cabrol, and Leclerq, , op. cit. (note 12), s.v. ‘Rouen’, xv (1), col. 128.Google Scholar

⁴² Ibid. s.v. ‘Ugium’, xv (2), cols. 2854-8 and fig. 11201.17.

⁴³ Chenet, G., La Céramique gallo-romaine d'Argonne du IV^e siècle et la Terre Sigilée décorée a a la Molette (Mâcon, 1941), 109–22,Google Scholar pls. xxv, xxv-xxviii. I am grateful to Ms Brenda Dickinson of Leeds University for this reference.

⁴⁴ Donovan, , op. cit. (note 26), 114–15.Google Scholar

⁴⁵ One further point could be added here: among the pottery finds from the Romano-British building excavated at Bourton-on-the-Water in Gloucestershire is a piece of fourth-century grey ware. Its decoration, rare in Britain, consists of three rows of a stamped pattern of a saltire within a circle, with a round ‘pellet’ in each quadrant. An exact parallel to this decoration has been found on a small vessel from a seventh-century Merovingian cemetery (Donovan, , op. cit. (note 26), 113–14Google Scholar and fig. 6). If the motif had been used by Christians as early as the fourth century, this could suggest a link between the pottery and the lead tanks found at Bourton and add weight to a proposed Christian presence somewhere in the large settlement there (see Webster, G., ‘Small towns without defences’, in Rodwell, W. and Rowley, T. (eds.), Small Towns of Roman Britain, Brit. Arch. Rep. 15 (Oxford, 1975), 59Google Scholar and fig. 4), if not in the actual building where the tanks were found in an abandoned state. Moreover, there is no reason to assume that, with the fall of the Empire, there was a break in the continuity of religious symbolism, although new motifs were introduced in the Byzantine period and some older ones from the Roman period declined in popularity.

⁴⁶ Gough, M., The Origins of Christian Art (London, 1973), pl. 17.Google Scholar

⁴⁷ Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘Nantes’, XII(1), fig. 8668.

⁴⁸ Ibid., s.v. ‘sarcophage’, XV (I), fig. 10800.

⁴⁹ S. A. Janthoudidis, ‘χϱιστιανιχαὶ ἐπιγϱαΦαὶ ἐχ Κϱήτης’, Αθηνᾱ, xv (1903), 4 95 97 163, 95; this inscription is illustrated in Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘Inscriptions grecques chrétiennes’, VII (1), col. 653.

⁵⁰ Ambrose, , De sacramentis, 3.2.Google Scholar

⁵¹ Conder, C. R. and Kitchener, H. H., The Survey of Western Palestine, 3 vols. (London, 1881–1883), III, 368.Google Scholar

⁵² The circles were evidently applied to the panels before they were cut to fit the circular base and the quarter-circle was lost in the joining process (Richmond, , op. cit. (note 34), 166)Google Scholar.

⁵³ Ramsay, W. M. and Bell, G. L., The Thousand and One Churches (London, 1909), 132, fig. 95.Google Scholar

⁵⁴ Guy, , op. cit. (note 2), 271.Google Scholar

⁵⁵ Toynbee, J. M. C., ‘Christianity in Roman Britain’, J. Brit. Arch. Ass. 3rd ser. xvi (1953), 23.Google Scholar So too Painter, K. S., ‘Villas and Christianity in Roman Britain’, Brit. Mus. Quarterly, xxxv (1971), 167Google Scholar.

⁵⁶ Toynbee, , op. cit. (note 55).Google Scholar

⁵⁷ Thomas, , op. cit. (note 3).Google Scholar

⁵⁸ Donovan, , op. cit. (note 26), 100.Google Scholar

⁵⁹ Hadman, J. and Upex, S., ‘Ashton, 1976’, Durobrivae, v (1977), 8Google Scholar; D. Miles, pers. comm.

⁶⁰ Guy, , op. cit. (note 33), 3.Google Scholar

⁶¹ Phillips, , op. cit. (note 24).Google Scholar

⁶² Kraay, , op. cit. (note 6), 219Google Scholar; West, , op. cit. (note 6), 79–81Google Scholar.

⁶³ Richmond, , op. cit. (note 34), 167.Google Scholar

⁶⁴ Guy, , op. cit. (note 2), 273.Google Scholar

⁶⁵ C. J. Guy, pers. comm.

⁶⁶ Richmond, , op. cit. (note 34), 169.Google Scholar

⁶⁷ Donovan, , op. cit. (note 16), 379.Google Scholar

⁶⁸ Curwen, , op. cit. (note 8), 156–7.Google Scholar

⁶⁹ Guy, , op. cit. (note 2), 274–5.Google Scholar

⁷⁰ Toynbee, , op. cit. (note 55), 15–16.Google Scholar

⁷¹ Frend, W. H. C., ‘Ecclesia Britannica: prelude or dead end?’, J. Eccles. Hist, xxx (1979), 129–44, 136.CrossRefGoogle Scholar

⁷² Thomas, , op. cit. (note 3), 226.Google Scholar

⁷³ West, , op. cit. (note 6), 78–9.Google Scholar

⁷⁴ Thomas, , op. cit. (note 3), 204–6 and fig. 39.Google Scholar

⁷⁵ Ibid., 221–5 and fig. 41.

⁷⁶ Ibid., 226.

⁷⁷ Ambrose, , De sacramentis, 2.16.Google Scholar

⁷⁸ Ibid., 1.19.

⁷⁹ Davies, J. G., The Architectural Setting of Baptism (London, 1962), 102Google Scholar; and the descent symbolized to the candidate his sacramental union with Christ in his death upon the cross as he was ‘buried therefore with him through baptism unto death’ (Romans 6:4).

⁸⁰ Rogers, C. F., Baptism and Christian Archaeology (Oxford, 1903), 314.Google Scholar

⁸¹ Thomas, , op. cit. (note 3), 220.Google Scholar

⁸² Dr Edward Yarnold, Campion Hall, Oxford, has drawn my attention to the fact that there were two fonts close together at Castel Seprio near Milan. It may be that we are looking here at endowments at a church, in the same way that church plate was duplicated. This does not necessarily mean that both fonts were used at the one time.

⁸³ Tertullian, , De baptismo, 17.Google Scholar

⁸⁴ Ambrose, , De sacramentis, 2.16.Google Scholar

⁸⁵ Ignatius, in the second century, says that it was not lawful for anyone but the bishop to baptize (Ad Smyrnam, 8), but Tertullian indicates that this function could be delegated to presbyters, deacons and even laymen at the beginning of the third century ( Tertullian, , De baptismo, 17).Google Scholar Even so, only the bishop could perform the imposition of hands, which was an integral part of the baptism ceremony. See Noakes, K. W., ‘From New Testament times until St Cyprian’, in Jones, C., Wain-wright, G. and Yarnold, E. (eds.), The Study of Liturgy (London, 1978), 80–94,Google Scholar for a discussion of the ritual before the fourth century.

⁸⁶ Chrysostom, John, Baptismal Instructions, 2.25Google Scholar (Wenger).

⁸⁷ Theodore of Mopsuestia, Baptismal Homilies, 3.14–15.Google Scholar

⁸⁸ Ambrose, , De sacramentis, 1.15Google Scholar; Cyril of Jerusalem, Catechesis, 3.3:Google Scholar that is, that this action brought the Holy Spirit down on it, giving it a supernatural effect (Yarnold, E. J., The Awe-Inspiring Rites of Initiation (Slough, 1971), 105, n. 28)Google Scholar.

⁸⁹ Baptism was normally performed only at Easter and Pentecost (Tertullian, , De baptismo, 19),Google Scholar though the sacrament was also administered at Epiphany in Cappadocia and perhaps elsewhere (Yarnold, E. J., ‘The fourth and fifth centuries’, in Jones, , Wainwright, and Yarnold, (eds.) op. cit. (note 85), 97)Google Scholar.

⁹⁰ West, , op. cit. (note 6), 78–9.Google Scholar

⁹¹ Thomas, , op. cit. (note 3), 226.Google Scholar

⁹² Ambrose, , De sacramentis, 3.5.Google Scholar

⁹³ This version from Canon Rock, ‘Celtic Spoons’, Arch. J. xxvi (1869), 42.Google Scholar Another, probably a later and amended version of this canon, reads: ‘neque pedes eorum lavandi suni a sacerdotibus vel clericis’ (Hefele, C. J., A History of the Christian Councils, from the Original Documents, to the Close of the Council of Nicea A.D. 325, trans. Clark, W. R. (Edinburgh, 1894), 157–8),Google Scholar which does make it appear that the Church in Spain no longer included this rite in its baptism ceremony. This does not, I feel, negate the argument for footwashing in the church of the West as a general practice in the fourth century.

⁹⁴ Ambrose, , De sacramentis, 3.4.Google Scholar

⁹⁵ Caesarius of Arles, Sermones, 64.2, 204.3.Google Scholar

⁹⁶ Augustine, , Ad Ianuarium, 2.33.Google Scholar

⁹⁷ Aphraates, , Homilies, 12Google Scholar (so Yarnold, E. J., ‘The ceremonies of initiation in the De sacramentis and De mysteriis of St Ambrose’, Studia Patristica, x (1970), 453–63,460)Google Scholar.

⁹⁸ The occurrence of footwashing in the baptismal ritual in the Church in Roman Britain could also explain the two rectangular structures associated with what is generally accepted as a fourth-century church at Silchester. The larger structure was probably the font, the smaller soakway for the pedilavium.

⁹⁹ Davies, , op. cit. (note 79), 26.Google Scholar

¹⁰⁰ Akeley, T. C., Christian Initiation in Spain, c. 300-1100 (London, 1967), 54.Google Scholar

¹⁰¹ Davies, , op. cit. (note 79), 26.Google Scholar

¹⁰² Cabrol, and Leclerq, , op. cit. (note 12),Google Scholar s.v. ‘baptistère’, II (1), col. 455 and fig. 1371.

¹⁰³ So Davies, , op. cit. (note 79), 26,Google Scholar for the Celtic examples. He does not accept this explanation for the eastern basins, however, since he believes ‘there is no trace of the pedilavium in the East’, and suggests they were ‘probably for the baptism of children to allow economy in the use of water’. Yarnold, while showing that footwashing was probably carried out in the East (Yarnold, , op. cit. (note 88), 27,Google Scholar n. 145), believes that all these small basins were subsidiary fonts, perhaps used for the baptism of children ‘who were too small to stand comfortably in the main font’ (ibid., 28). I would, nevertheless, question the need for permanent fonts for the baptism of children in the fourth century.

¹⁰⁴ Private baptism lasted as late as the sixth century. It was forbidden, except in cases of necessity, in 527 at the Council of Dovin in Armenia (canon 16) (Rogers, , op. cit. (note 80), 315)Google Scholar.

¹⁰⁵ Davies, , op. cit. (note 79), 106–7,Google Scholar says separate baptisteries were generally found only in episcopal cities until the time of Augustine.

¹⁰⁶ See above, note 89.

¹⁰⁷ See above for a discussion of dating evidence. 108 On the other hand, Ambrose believed it was because of the large numbers of candidates that the Church at Rome discontinued the pedilavium (De sacramentis, 3.5). However, there is no evidence to suggest that Ambrose's influence went westwards; moreover, his episcopacy was several decades after Julian's reign and any pagan revival in Britain.

¹⁰⁹ Guy, , op. cit. (note 2), 275.Google Scholar

¹¹⁰ Ambrose, , De sacramentis, 3.4.Google Scholar

¹¹¹ So Thomas, , op. cit. (note 3), 221–5.Google Scholar

¹¹² Ibid., 213-17.