Very few classical temples are known in Britain. Colchester’s temple to the deified Claudius, rebuilt after the Boudican revolt, is the largest yet discovered, but only its foundations survive. Not only was the site adapted for use as a Norman castle, but the temple seems to have been substantially remodelled in antiquity, possibly into a church. Like the slightly later Fishbourne Palace, the Colchester temple’s appearance in Britain so early was a revolutionary innovation. It was a massive structure, and the remains of its podium beneath the castle still stand almost 3.5 m (11 ft 6 in) above the Roman ground level. Its overall size must have been at least 32 x 24 m (105 x 79 ft). Based on these proportions, it has been estimated that the facade had two rows of eight columns, with another sixteen around the cella, where the cult statue and valuables were stored. The temple was seen by the Boudican rebels as the ultimate symbol of Roman oppression, and, appropriately enough, it became the site of the last stand of the colonists c. 60-61. Following the revolt the temple must have been rebuilt, but did not survive in its original form to the end of the period.
As the principal, or at least original, centre of the imperial cult, the size of the Colchester temple is only to be expected. Something similar must have stood in London, but nothing has ever been found. Such temples played an extremely important part in the annual cycle of official religious festivals. At St Albans, two classical temples of indeterminate form stood on the south side of the forum, but the five-odd other temples known in the town were all Romano-Celtic in design - even the prominent temple in its large precinct next to the theatre. At Wroxeter, a small classical-style temple seems to have been squeezed into a street frontage after a late second century fire had destroyed a house formerly on
The site.17 But this is an unusual find. Silchester, where most of the masonry structures are known, had no classical temple at all.
The Gorgon pediment sculpture from the temple at Bath. Minerva wore a Gorgon medallion and was conflated with Sulis, the deity of the sacred spring. Late first century.
The only Romano-British classical temple known in any detail is the temple of Sulis-Minerva at Bath. Although the building lies deeply buried, with large parts inaccessible, the recovery of pediment sculpture, significant parts of its columns and other architectural details make its original appearance virtually certain [229]. It undoubtedly went up in the Flavian period as a fairly standard tetra-style temple, although the hybrid elements of the pediment mark it out as a building with a very idiosyncratic flavour and iconography. By the fourth century it seems to have been significantly redesigned. Flanking shrines were built on either side of the temple steps, and some sort of covered ambulatory may have been built around the cella, creating a structure that had elements of both classical and Romano-Celtic design.
Fragments of other possible classical temples have been found elsewhere, but usually in a context that makes any sort of visual restoration impossible. A small fragment of a massive inscription in Winchester of Antonine date can only have come from a major classical temple or administrative building. The temple of Diana recorded on an inscription found at Caerleon is a good candidate for a classical temple, but no structural remains are known. A large section of entablature, very possibly from a classical temple, was found built into the late Roman southwest gate at Lincoln, acting as a chamfered plinth block in one of the flanking towers. Evidently the temple, if that is what it was, had been demolished for a long enough period of time for an odd block to be reused by persons unknown who neither knew nor cared where it had come from.
The temple of Sulis-Minerva at Bath as it might have appeared in its original form in the late first century, looking east across the temple precinct.
THE CONTRACT
One of the key characteristics of Romano-British pagan cult activity was the ancient notion of contract. Normally we see the end result of this pact on stone altars in which the abbreviation ‘VSLM’ appears. The letters stand for votum soluit libens merito and means ‘he willingly and gratefully fulfilled the vow’, where ‘he’ (or she) is the person responsible for offering the altar. The dedicator had requested a service from the god or goddess, and in return an altar and offerings were promised in exchange. The only context in which we ever usually find evidence for the vows is in the ‘curse’ tablets found at Bath and Uley, and at isolated, scattered finds. The only traces we normally have of messages to gods made on wooden writing leaves are the bronze seal-boxes that secured any petition made and kept them private. The many seal-boxes found at Great Walsingham (Norfolk) are thought to be evidence for an as yet undiscovered shrine.
What these ‘curse’ tablets tell us about individuals has already been discussed in Chapter 9, but in religious terms they record the process of the transaction through carefully worded formulae. At Uley, for instance, an unnamed devotee addresses ‘the holy god Mercury’ to complain about ill-treatment, and requests that the god prevent those responsible from standing, sitting, eating, or drinking, and render them unable to remove the curse unless their blood is spilled.18 To make sure all potential candidates are covered, the stock phrase si servus, si liber, si mascel, si femina (‘whether slave, free, man, or woman’) is added. Stone altars usually survive, but offerings could be made and burnt on wooden or earthen altars (both are attested in ancient literature), or buried in pits. Altars were deposited in almost any sacred context, and have been found in water shrines, inside or near temples, at remote open-air rural shrines, or in what now seems to be total isolation. Although religious dedications could be made at any time or any place, the sacredness of a site and the date in the calendar were very important. Much religious ritual was performed according to an annual cycle of special occasions. Usually some sort of procession came first, depending on the context. In towns, the theatres and temple precincts were normally connected by roads, creating a standardized route to be followed for official ceremonies. Private ceremonies may have involved a procession around the boundaries of a farmstead, a villa estate, or an urban plot.
The offerings preferably involved the sacrifice of a living animal, together with wine libations and other foods, followed by an examination of the entrails in search of good or bad omens. Not surprisingly, the physical evidence for sacrifice is almost impossible to identify beyond the recovery of animal bones at temple sites, where the presence of a particular species may reflect the deity being worshipped. Mercury’s principal associates were the goat, ram and cockerel. Appropriately, of the quarter of a million bones analyzed from his shrine at Uley, the vast majority were from these animals, and many must have come from sacrificial animals sold to pilgrims in the temple complex.
The amphitheatre (Maumbury Rings) at Durnovaria, just outside the town walls. It dates to c. 70-80 when the Romans adapted a Neolithic henge monument by digging out the middle as an arena. This perpetuated the use of the site for religious reasons since, like all amphitheatres, it was used for games and entertainments associated with religious festivals.
Religious activity had to be observed strictly according to ritual to be valid.19 It was essentially a superstitious approach. If something unpleasant or disastrous occurred, then it would be concluded that the rituals had not been performed properly. Serving as a priest in the Roman world, and carrying out those duties exactly as prescribed, was an integral part of a man’s social responsibilities. Priests were not full-time members of a separate caste. The idea of a professional priesthood represented something of a threat to the Roman government, since it diluted the emperor’s authority. The emperor was the chief priest, or pontifex maximus. Senators served as priests in the course of their other duties, and regarded the position as an honour. Maintaining religious traditions was considered beneficial to the welfare of the state and the community in general. Throughout the Roman Empire, provincial governors and civic officials had responsibilities for observing religious festivals. Other men of lesser rank included priestly duties in their daily routines. Regalia formed an important part of a priest’s ceremonial appearance. Priests wore headdresses made of bands or chains, or diadems, and carried staffs or sceptres capped with heads that either represented emperors or deities.
Lucius Marcius Memor was a haruspex, or soothsayer, at Bath, the only one known by name in Britain [231].20 The haruspex, who examined a sacrificial victim’s entrails, inherited a tradition stretching back to Etruscan times and was a familiar part of Roman religion. Gaius Calpurnius Receptus, sacerdos deae Sulis (‘priest of the goddess Sulis’), was also at Bath, and another two priests are known on Hadrian’s Wall. 21 Marcus Aurelius Lunaris, probably a freedman trader, served as a sevir Augustalis at both York and Lincoln, meaning that he was a board member of the imperial cult at both colonies.22 Another trader, Marcus Verecundus Diogenes, was also a sevir of the colony of York. 23 At Greetland (West Yorkshire) in the year 208, Titus Aurelius Aurelianus described himself as a magister sacrorum (‘master of the sacred ceremonies’).24 Since his altar commemorated a private family offering to Victoria Brigantia and the imperial spirits of Septimius Severus and Caracalla, he may have adopted this rather grandiose title in his capacity as master of his own house, a reminder that the senior man in any context had religious responsibilities to fulfil.
Statue base from the temple precinct at Bath, recording Lucius Marcius Memor, haruspex (soothsayer).
At the fourth century healing cult of Nodons, at Lydney, two rather more unusual religious posts crop up. Titus Flavius Senilis, recorded in an inscription on a now-lost mosaic from the temple, was described as a ‘PR REL’, for which the only plausible expansion is praepositus religionis, or ‘director of the cult’, though the term praepositus in this context is certainly unusual.25 The same inscription mentions Victorinus the interpretiante, or ‘dream interpreter’. The interpretation of dreams as spiritual messages was well established in antiquity, dating back into pharaonic times in Egypt. It was commonplace for people to intensify their sleep experiences by drinking themselves into oblivion first. The physical
Structure of the temple at Lydney seems to have been designed with recesses for pilgrims so that they could dream their drunken dreams in the vicinity of the god, and wake in the day to have them ‘read’ by the resident expert. At Bath, a man whose name is lost but who called himself ‘son of Novantius’, erected his monument, presumably to the goddess, ex visu, ‘as the result of a vision’.26
TEMPLES AND SACRED PLACES
Most prehistoric places were venerated for their natural features, and structures were not integral or necessary to the concept of sanctity. The only pre-conquest buildings identifiable as temples known in Britain were in the southeast, such as the timber circular ‘tower’ cella and enclosure at Hayling Island, in use by the first century BC [232]. They seem to have been yet another dimension of Continental culture already being absorbed from across the Channel before 43.27 Wanborough (Surrey) had a circular temple by the late first century, but it lay on top of possible earlier structures. 28 The vast numbers of Iron Age coins found on the site make it likely that there had been some sort of pre-conquest temple structure on the site. Wanborough is also an early instance of coins being used as votive deposits, a characteristically Roman habit. Hayling Island’s pre-conquest temple was replaced with a Roman version in stone on a similar but much larger plan. The new version had a porch, and was surrounded by a masonry precinct wall and rooms. By the mid-second century, Wanborough had a conventional Romano-Celtic temple. Sacred sites that remained in their natural state were also liable to development. The most conspicuous and exciting example of adaptation was at Bath. Others, like Coventina’s Well at Carrawburgh, were left more or less as they were.
232. Hayling Island (Hampshire).
Reconstructed view of the circular temple at Hayling Island as it might have appeared in the first century. The Roman development of the site represented continuity of use of a much older shrine.
Temples in Roman Britain fall mostly into two categories: classical and the much more common Romano-Celtic. The very few classical temples known were in towns. Romano-Celtic temples were built in large towns, small towns, rural settlements, and in total isolation. In all cases, the temple was the home of the god or goddess. A representation of the deity was stored within the cella. Cult activity mainly took place outside in the temple precinct (temenos). It was not unusual for several temples, especially Romano-Celtic temples, to be sited within the temenos, each dedicated to a different deity. Outside the temples stood altars on which ceremonial sacrifices took place, and over which sacred incantations were spoken. The temples provided an attraction to pilgrims and travellers, and formed an integral part of the services provided in towns. Godmanchester (Cambridgeshire), for example, had a temple next door to a building interpreted as the mansio (inn).
In some instances, temple-building was ambitious, or at any rate unusual. The shrine of Apollo Cunomaglos at Nettleton [235] probably had I ron Age origins. At its climax, the temple was remodelled as an innovative and ambitious octagonal structure, based on the Romano-Celtic temple concept of internal cella and surrounding ambulatory. The octagonal form was apparently beyond the architect’s skills, and he failed to recognize that the central lantern of the cella would place huge pressure on the external
Walls. Consequently, no provision for buttressing was made, and the temple eventually suffered partial collapse. Equally unusual, but far more successful, was the probable temple to Victory at Carron, near Falkirk, on the Antonine Wall [236]. Built as a beehive structure with dressed stone, the design allowed the weight of the domed roof to be carried down through the walls. The building was so well executed that it survived intact until the eighteenth century, the only Romano-British structure to do so, until it was dismantled to provide stone for a dam.
The octagonal shrine of Apollo Cunomaglos as it may have appeared before collapsing in the mid-fourth century. The surrounding settlement survived as a centre of light industry, including pewter manufacture. It has been suggested that the ruins were even adapted into an early church, but this has not been proved. The temple and its attendant settlement lay on the Fosse Way between Cirencester and Bath, and in its heyday clearly benefited from and relied on passing trade.
This remarkable ‘beehive’ temple was probably built by the Roman army as a temple to Victory. Mid - second century.
In contrast, ritual in the congregational mystery religions of the Roman Empire took place mainly indoors. The temples of Mithras were based on the basilican form of nave and aisles - a form later adapted for use as Christian churches. Not many early Christian churches are known in Britain, although Silchester’s fourth-century ‘church’ beside the civic basilica is one of the most convincing, as is the cemetery ‘church’ at Colchester [237], with its ground plan of eastern apse, nave and aisles.29
This ‘church’ has been identified as such because of the ground plan and clustering of late Roman east-west graves around it. Built about 320-40, with the apse added c. 380.
The inscriptions from Roman Britain that record temples usually commemorate an individual’s endowment of the building or an associated structure, either in a personal or professional capacity. Most come from the north, even if the donor was apparently civilian. At York in 221, the trader Lucius Viducius Placidus built an arch, evidently for a temple precinct, as did Trenico, probably for the precinct of the temple at Ancaster to Viridios, who is named on the slab.30 At Caerleon, Titus Flavius Postumius, legate of (presumably) the II legion, restored a temple of Diana around the mid-third century.31 Cooperative efforts are also recorded, such as the building of a temple to the Mother Goddesses somewhere near milecastle 19 on Hadrian’s Wall by a vexillation from the First Cohort of Vardullians. 32 The inscription from Chichester recording the client king, Togidubnus, also records the dedication of a temple to Neptune and Minerva by a guild of smiths. 33 And in London, an imperial freedman called Aquilinus probably restored a temple to Jupiter Optimus Maximus, with assistance from three other individuals, probably in the third century.34
The renewed interest in paganism in Roman Britain in the late third and fourth centuries, despite the progressive outlawing of pagan practices, was manifested in the countryside where a number of new temples were
Established, usually as fairly isolated buildings. Urban temples, however, generally fell out of use. Lamyatt Beacon (Somerset), begun in the late third century, stood on a ridge with epic views across the Fosse Way to the west. Brean Down, not far away, was begun in the early fourth century. At Maiden Castle, within the old Iron Age fortress, a new Romano-Celtic temple was built on a virgin site after 367, dated by a hoard found under its tessellated floor. Nearby at South Perrott (Dorset), coins of the late third and fourth centuries were deliberately inserted in pits around what is now known to have been a Neolithic or Bronze Age ditched enclosure, probably reflecting a revival of veneration for ancient monuments in the landscape. Similar deposits have been found at many other prehistoric monuments. However, it is extremely unlikely that those who buried the coins had any idea at all of what those monuments had once been used for.
The temple of Nodons at Lydney was originally built in the late third century on a site that must have already been sacred to Nodons, but which had presumably remained undeveloped. Many of these new rural temples were in regions where the most luxurious villas were being developed, often with iconography on their mosaic floors reflecting an intense interest in pagan myth and cult activities. These are characteristically difficult to interpret, or to connect with temples. The triconch hall at Littlecote with its distinctive Orpheus mosaic has been interpreted as the meeting place of an Orphic cult, or as an elaborate summer dining room. Mosaics are virtually unknown in Romano-British temples themselves.
Altars and other votive goods were often set up in the open air. Outside the temple of Mithras, at Carrawburgh, a small stone bench and an altar marked an outdoor shrine to the local Genius and the Nymphs. A few metres to the north, the shrine of Coventina [ 238] was no more than a stone-lined revetment around the site where her spring rose. Making offerings to a local deity was a way of harnessing that deity’s power to Roman ends. Failing to do so might, in the Roman superstitious mind, have risked its wrath. In such a context, building a physical temple was unnecessary, although there is no doubt that one was built at Bath. At Carrawburgh, the shrine to Coventina was left as it was, though now she was offered gifts and portrayed in the Roman way.
238. Carrawburgh (Northumberland).
Relief from the spring at Carrawburgh (Brocolitia). Dedicated to the nymph Coventina by Titus D[...] Cosconianus, prefect commanding the First Cohort of Batavians.
This explains the recently-discovered shrine of Senua, near Baldock, where a quantity of votive gifts, including gold and silver plaques, jewelry, and material deposited by the Spaniard, Servandus, were thrown into a pool. There were probably attendant buildings around the pool to accommodate pilgrims and service the cult. Senua, who is otherwise totally unknown, was probably the Celtic goddess of the water in the pool. Despite the fact that she had been Romanized by being depicted as Minerva, the
Centre of the cult remained a pool, and not a temple.
Incidental places could become a setting for an act of veneration. On Bollihope Common, near Stanhope (Durham), Gaius Tetius Veturius Micianus, prefect of the Sebosian cavalry wing, hijacked an altar that had already been dedicated to the imperial spirits. He recarved it with a dedication to Silvanus Invictus (‘Undefeated Silvanus’), in gratitude for the god’s aid in helping him to capture an evasive wild boar. 35 At nearby Eastgate, the prefect, Aurelius Quirinus, made another dedication to Silvanus [239]. To the south, on Scargill Moor, an open-air shrine to Vinotonus Silvanus stood by a stream where several soldiers left other altars in this remote and windswept spot.36 At Custom Scrubs, Bisley (Gloucestershire), a pair of gabled reliefs depicts Mars Olludius and Romulus, respectively. There is no recorded structural context for either of these reliefs, which were probably displayed in an open-air shrine. Not far from the fort at Risingham, a relief depicting an unnamed warrior god, known locally as ‘Rob of Risingham’, is still partly visible in situ. All of these altars and reliefs are typical of countless roadside or rural shrines that were scattered across the countryside in every province of the Empire, to which passers-by, travellers and hunters made offerings as a matter of course. In the Christian era, the more obvious and accessible examples were susceptible to destruction, and even ‘Rob of Risingham’ was partly destroyed in the nineteenth century by the landowner who was sick of visitors. However, anyone passing through Catholic countries in Europe and Latin America will have become familiar with Christian wayside shrines, which preserve a much older pagan tradition.
Altar dedicated to Silvanus by Aurelius Quirinus, commanding officer of the First Cohort of Lingonians at Lanchester under Gordian III (238-44).
World History
