Temples

Temple structures are usually recognised in the archaeological record through their distinctive building plan, either being of classical plan or, more usually in Britain, of Romano-Celtic design (Wilkes 1996:1). This Romano-Celtic design of temple consisted of a square chamber (cella) surrounded by an ambulatory and this could also be set within a larger precinct (temenos). Despite this, there is still uncertainty about the identification of some buildings as temples in Britain. Where only small areas of the buildings have been uncovered, the interpretation of the structure as a temple is often problematic. At Gloucester, for example, the remains at Westgate Street were originally interpreted as the edge of a bathhouse (Heighway and Garrod 1980), but Hurst (1999b: 155-7) suggests that a peribolos (court enclosed by a wall) of a temple might be more likely. At Cirencester, excavations within insula VI opposite the basilica uncovered an area of courtyard and section of the portico of a building of monumental nature that have led to suggestions that it may have been the temenos of a temple, although little else is known (Holbrook 1998: 139-40). In the case of small towns, temples often seem to have been the only public building and they sometimes had a central position, as at Elms Farm, Heybridge, in Essex (Atkinson and Preston 1998), and Westhawk Farm in Kent (Booth 2001). This central location and surrounding open spaces indicate that the temples were perhaps involved in market and administration activities.

Within Roman London a number of large monumental complexes are known. Their functions remain enigmatic but perhaps included some kind of religious role without the buildings being wholly temple complexes. The complex described as an 'Allectan Palace’, by the side of the Thames, excavated on small sites including Peter’s Hill and Sunlight Wharf, was built in the later third century. The fact that the date of its construction in a. d. 294, dated by dendrochronology, coincides with the reign of Allectus has led to suggestions that it was a palace built by Allectus; however, there is no definite proof of this (T. Williams 1993: 28). It reused masonry apparently from an earlier complex on the site and baths here make a religious interpretation possible (T. Williams 1993: 26-32). Further masonry that apparently came from these buildings was found within the late riverside wall. This stonework displayed religious features, including depictions of gods that had come from religious monuments that had been part of the complex (C. Hill, Millett, and Blagg 1980: 125-32). It is uncertain whether the new buildings continued any of the functions of the earlier ones. Across the Thames at Southwark there was another large complex of uncertain function, which included a number of large wings of rooms and a bathhouse of comparable size to that at Huggin Hill (Yule 2005: 50-72). Some sort of religious or military role is a possibility here. Relatively few sites in urban contexts have produced many obviously religious artefacts such as statuettes and regalia. This contrasts with some rural sites such as Uley (Woodward and Leach 1993) and Hayling Island (King and Soffe 2001), and it could be the result of post-Roman reuse and disturbance of the urban sites, or perhaps of the differing rituals that occurred in the buildings.

Temples are also useful when one is considering movement around, to, and from towns because they would have played an important role in religious ceremonies and festivals (Esmonde Cleary 2005); they may well have attracted people to the towns from long distances. Some temples would also have had restrictions of access both in terms of time (of day and year) and type of person, reflecting aspects such as class (Stambaugh 1978: 574-80). This will have intensified the experience of entering the temple and worshipping within. Whereas other aspects of the landscape would have been imbued with meaning, including natural features, temples were an important way in which the religious landscape was created and negotiated. Roman temples in Britain were also apparently sometimes located on sites of pre-Roman religious importance (A. Woodward 1992:17-30), indicating a longer-term continuity of activity (ibid.: 63).

Despite the presence of temples and shrines within towns, it must also be acknowledged that there were other forms of religious expression within the urban centres (see Chapters 3 and 4; cf. Fulford 2001). Religious activity took place at other public buildings, including theatres, amphitheatres, the forum-basilica, and bathhouses, and towns also formed part of the wider ritualised landscape. Temples and their precincts encouraged many diverse public activities, including meetings, business transactions, and performances (Anderson 1997: 243; Perring 1991b: 280; Stambaugh 1978). In some cases there are traces of activity beneath the temples, as at Canterbury (Frere 1977:423) and Verulamium (Lowther 1937), and, despite the caution required in assuming a continuation of religious activity, it is possible that these were the locations of pre-Roman shrines representing continuity in the religious landscape.

A number of temples display structural alterations in the late Roman period indicating a change or continuation in use (Table 5.4). Of the thirty-eight definite temples known within the towns, fifteen buildings had at least parts of the structures standing into the later fourth and fifth centuries. A further five temples may have been standing, though there is insufficient evidence to be certain, and only three definitely appear to have been demolished by the fourth century. For a further fifteen known temples there is not enough evidence for an analysis of their structural condition in the late Roman period to be possible.

5.5.1 Caerwent

The temple at Caerwent, which was not built until around a. d. 330, was maintained throughout the fourth century, and into the fifth, with evidence of a number of late alterations. These included the addition of a range of rooms to the inner side of the entrance hall, suggesting that more space was needed for the activities taking place in the hall (Brewer 1993: 59). A coin of Valentinian (a. d. 364-75) was found beneath repair work to the foundations of one of the pilasters in the entrance hall (Brewer 1990: 79; 1993: 59). There have been various interpretations regarding the role of this entrance hall and rear range of rooms. Reece (personal communication) and Knight (1996: 36) have argued that the hall may have taken on the role of the basilica, after that had decayed, because it would have been more convenient and economical to maintain. Other interpretations have

Included priests’ quarters and shops selling religious votive gifts and souvenirs (Wacher 1995: 386). The hall appears to have remained in use into the fifth century, although there is no definite evidence of its use and, as discussed, there are reasons to suggest that at least parts of the forum-basilica remained standing to a contemporary date.

5.5.2 Colchester

At the Balkerne Lane temple in Colchester there is evidence of differential robbing activities. The ambulatory of the temple was completely robbed, including its foundations, in the late Roman period (P. Crummy 1984:125), but the foundations of the cella survived to the height of the latest Roman layers, indicating that it was only after abandonment of the building that these walls were demolished or had collapsed. It would appear that the cella stood in isolation in the fourth century, which might indicate a different function; the excavator suggested a church (ibid.) but its continuation as a temple or for another use is also quite possible.

Little is known of the Temple of Claudius in Colchester but some of its walls are visible within the cellars of the medieval castle. Studying the known evidence and plan, mainly from excavations in the 1930s, Drury (1984) suggested that there was evidence of a 2-m-thick wall built across the front of the temple in the fourth century and also for the demolition of the temple facade. He suggested that this created a long and narrow space with an apse, a church being formed by the conversion of the building. The latest coins on the site were of Valentinian II (a. d. 382-93) and Theodosius I (a. d. 379-95), and pottery found dated up to a. d. 360-70. Excavations in 1996 beneath the castle in the supposed location of the wall, however, failed to support these earlier findings, arguing that this had merely been a hypothetical projection (P. Crummy 1997:120), but the building appears to have remained standing. This need not necessarily mean that the function of the building did not change in the later Roman period and, though clearly not enough evidence exists, a piece of pottery with a chi-rho symbol was found there in the 1996 excavations.

5.5.3 Verulamium

Another example of a temple where conversion to a church has been suggested is the insula XVI temple in the centre of Verulamium near the theatre, although the excavations of this structure took place in the 1930s. Lowther (1937: 33-4) suggested that at around a. d. 400, indicated by coins of the House of Theodosius, a new gateway to the building was constructed in the centre of the colonnade on the west side of the temple. This western gateway appeared to replace the earlier eastern entrance (ibid.) and therefore might indicate a change in orientation of the building. Although this need not equate with its conversion to a church, it would indicate that the building remained in some kind of use into the fifth century and possibly beyond.

5.5.4 Canterbury

At Canterbury, the small excavated areas of the central temple precinct indicate that demolition and robbing of the precinct portico and levelling of at least some internal buildings took place in the fourth century; the area was covered by a new courtyard surface on which were many fourth-century coins (see Section 6.2.1; unpublished excavations, P. Bennett personal communication; Frere 1977: 424). The new surface indicates that the area continued to be important within the town centre. The other temples known at Canterbury are only partially excavated and many have been badly disturbed. The temple at Gas Lane has coins dating to a. d. 330-40 in disturbed layers (Bennett, Frere, and Stow 1982:44), but there is nothing to indicate that it was demolished and it may have remained standing to a later date. The Burgate Street temple has late-fourth-century pottery in its demolition layer (Frere and Stow 1983: 47), but this need not necessarily indicate the date at which the building was destroyed because the act of demolition can disturb earlier layers.

5.5.5  London

At London, masonry of the riverside temple complex was incorporated into later buildings on the site and elsewhere, indicating that it had been destroyed by the end of the third century (T. Williams 1993: ii, 27). The monumental nature of the structures that were then built on the site may, however, suggest a continuation rather than change in use (ibid.: 28-9). At the temple of Mithras there are alterations to the structure in the fourth century with the removal of the columns that had separated the nave from the aisles, apparently creating a larger open space (Shepherd 1998: 84). This appears to have coincided with the burial of many religious sculptures within the building. The definite function of the structure is uncertain71 but it remained standing into at least the sixth century (ibid.: 97). Internal alterations to the building in the late fourth century, creating a more open space, suggest that many people were still using it.

5.5.6  Other towns

Late phases of temples from the other towns in the study have not survived well. These include the temples at Caistor-by-Norwich (Atkinson 1930; Gurney 1986), Wroxeter (Bushe-Fox 1914), and Leicester (Wilson 1970:286), although there were some late-fourth-century coins from this latter site. At Winchester, the temple was demolished in the third century (Biddle 1975: 299). No temples have been identified with certainty at Cirencester, although the enigmatic monumental building in insula VI, of which only a small area has been excavated (Holbrook 1998: 135-8), may have been one. Results from this site, including well-worn coins of ca. a. d. 400, indicate that it remained standing into at least the fifth century (ibid.: 135-8). At Gloucester, there is an equally problematic building excavated at 63-71 Northgate Street that, it has been argued, functioned as a temple. The excavations indicated floor resurfacings that sealed coins of the fourth century and changes to the internal walls, these seemingly being converted into arcades or colonnades in a later period (Hurst 1972: 65). This would demonstrate use well into the fourth century and perhaps beyond.72

That the use of temple buildings in the Roman period was variable, with many activities taking place within and around them (Stambaugh 1978), makes studying their function in the late Roman period more complex. It is perhaps overly simplistic to rely on historical documents such as the Theodosian Code, which banned the use of temples for pagan religion and ordered their preservation for alternative use, to date the end of use of the buildings (see Section 3.2.3 for these code entries). There are well-known examples in Rome where temples were converted into churches (Webb 2000), but there is very limited evidence in Britain for such use of these structures. Examples include the temple of

Fortuna Virilis, which became the church of Santa Maria ad Gradellis between a. d. 872 and 88o, and the church of San Nicola, which was constructed in the Forum Boarium out of the parts of three adjoining temples (Hansen 2003: 182). It is not easy to determine similar processes in Britain, and the location of temples within the townscape and wider ritualised landscapes may also be an important reason why some of the sites remained in use. It is important not to see evidence of structural decay as always indicating the end of the value of a site.