SYMBOLS OF POWER AND NATURE: THE ORPHEUS MOSAICS OF FOURTH CENTURY BRITAIN AND THEIR ARCHITECTURAL CONTEXTS
The aim of this paper is to demonstrate the importance of studying
mosaics within the social/architectural context. As Preston Blier (1987: 1)
has pointed out, architecture is invariably anthropocentric, being bound up
with human activity, experience, and expression. Mosaics, as integral features of architectural design, must also be seen as socially constructed. The
approach to be employed here will involve the identification and location
of those sites possessing Orpheus mosaics, and an analysis of the nature of
representation. A part of this analysis will consider similar representations
in other contexts and media, in order to provide new insight into’ the possible significance of Orpheus within fourth century Britain. The final part
of the discussion will concentrate on the social/architectural contexts of
the pavements, and an attempt will be made to explain the evidence in the
light of this.
THE DATA —-
Before carrying out any kind of analysis it is necessary to identify and
locate representations of Orpheus on floor mosaics, and to place them
within some kind of chronological framework. It will be useful to divide the 106
… ! i
G 0 ‘–
. , ‘\ II
Figure B.1. Rooms with mosaics at Withington villa
(corresponding mosaics in Figure B. 2) (after L ysons 1B17).
40 feet SYMBOLS OF POWER AND NATURE: ORPHEUS MOSAICS 107
Figure 8.2. The mosaics at Withington villa (after Lysons 1817). 108 S. SCOTT
pavements into two groups: definite examples, and possible examples.
Within the first category there are eight pavements: Barton Farm (Buckman and Newmarch 1850); Withington (Lysons 1817) (Figures 8.1 and 8.2);
Woodchester (Lysons 1817); Newton St. Loe (Nichols 1838); Littlecote
Park (Hoare 1819) (Figure 8.3); Winterton (Stead 1976) (Figure 8.4); Horkstow (Hinks 1933), and Brading (Price 1881) (Figure 8.5). Within the
second category there are a further seven pavements: Dyer Street
(Beecham 1886); Whatley;} Wellow;2 Pit Meads (Hoare 1819); two pavements from Caerwent (Toynbee 1964: 266), and some fragments from
Bishopstone.3 The distribution of these pavements is summarised in Figure
8.6. The main aspects of the data are summarised in Figure 8.7.
In considering art forms in the archaeological record it is necessary to consider not only what is represented, but also how it is represented. As Boas
notes (1955: 13):
It is essential to bear in mind the twofold source of artistic effect,
the one based on form alone, the other on idea associated with
form. Otherwise the theory of art will be one-sided.
With regard to Romano-British Orpheus mosaics, it will be important to
assess the ‘meaning’ of Orpheus in the light of classical mythology, while
also paying attention to the formal relationships within the mosaics, and
the effect of form on meaning.
In mythology, Orpheus was the greatest singer and musician conceived
by the Greeks. He was the son of the Muse Calliope, by either a King of
Thrace or Apollo. Apollo gave him a lute with which he was able to charm
wild beasts and make rocks and trees move. Jesnick (1989: 10) has noted
that the character of the Orpheus image is defined by its difference from
other scenes in the animal genre:
for the singer achieves by his art what otherwise takes great
physical courage and skill. Men have only a tenuous hold over the
captured beasts, who, always seeking to escape, eventually kill, or
are killed. Orpheus stills this endless cycle.
The animals traditionally found on the Orpheus scene are those which
exhibit types of behaviour which make them difficult to handle or capture.
The animals represent both the negative and the bountiful aspects of SYMBOLS OF POWER AND NATIJRE: ORPHEUS MOSAICS 109
Figure 8.3. Mosaics at Littlecote Park (after Venue 1730). 110 S.SCOlT
Figure 8.4. Mosaic at Winterton (after W. Fowler 1796).
nature, which Orpheus is able to control and hold in balance. The lion and
the leopard are those animals most frequently represented in the RomanoBritish pavements, as is the case throughout the empire.
The majority of the Romano-British Orpheus mosaics employ the following scheme (Figure 8.8):
Orpheus is usually found at the centre of the pavement, surrounded by
animals and birds. The central position of Orpheus emphasises his power,
while the surrounding concentric circles reflect the nature of this power; SYMBOL’) OF POWER AND NATURE: ORPHEUS MOSAlCS III
.. ::: .
– ••••• ‘o ••••
Figure 8.5. Mosaic at Brading (after Price 1881).
the animals are ‘contained’ by his charms. Another important feature of this
design is that it can be viewed from any angle and still be understood. It is
also possible to comprehend the central figure as someone possessing power
without having to understand who he is, or the mythological background.
If we are to further understand the significance of Orpheus within
Romano-British mosaics, it will be useful to assess the evidence for similar
forms of representation in other contexts and media.
REPRESENTATIONS IN OTHER CONTEXTS —-
There are a number of representations in Britain which share physical attributes with Orpheus i.e. they are wearing similar clothing and are accompanied by animals. For example, at London (Merrifield 1986); Chedworth
(Toynbee 1962: 156); Bisley (Clifford 1938); Upton St. Leonards (Rawes
1977); Box (Toynbee 1964: 179); Wilsford (Toynbee 1964: 179); Nettleton 112 S.SCOTI
(Wedlake 1982); Thruxton (Ingram 1849), and Lydney (Wheeler 1932).
1bree sculptures from London, tenned ‘hunter gods’, of mid-fourth
century date, are dressed in short tunics, a cap similar to that worn by
Orpheus, and carry a short sword. At least two of the examples are accompanied by a dog and a deer. Merrifield (1986: 87) suggests that these
figures probably came from a substantial Roman building which might
have been a temple complex, the meeting place of a religious guild, or a
residence with private shrines. Merrifield (ibid.: 89) proposes that these
figures, and others with similar characteristics from the sites mentioned
above, are the result of highly sophisticated and constructive religious
thought. In support of such an argument are the parallels between the
hunter god and Apollo and Orpheus. There is a relief from Ribchester,
identified by its inscription as Apollo Maponus, in which the composite
deity is represented with a quiver on his back, a lyre by his side, and apparently wearing a cap. It is likely that Apollo acquired this head-dress by a
double association; with Orpheus through their common attribute, the
lyre, and with Mithras through his solar connections (Merrifield ibid.).
Hunter gods from the north and south-west have also been identified with
Silvanus, god of the woodlands and wildlife.
To summarise, it seems that there was a cult figure in late Roman
London, almost certainly under official patronage (ibid.: 87), who appears
to be a conflation of Apollo, an oriental mystery god, and a native deity
representing nature. It suggests someone in authority in London with a
strong interest in comparative religion and the initiative to develop a new
syncretic cult. This cult seems to have been popular in southern and
central south-western England. Returning to the Orpheus mosaics, it can
perhaps be suggested that his popularity, particularly in the central southwest, had something to do with his conflation with the hunter god
identified above (Henig 1986: 17). In addition to the shared physical
similarities, there are also the shared connections with power and nature.
It is interesting at this point to consider the interpretation of the Littlecote pavement proposed by Walters (1982). Walters has suggested that
Orpheus acts as a link between Apollo, of whom he was a priest, and
Dionysus, the principal deity of the cult alluded to in the mosaic. Weare
meant to see Apollo and his priest Orpheus. The same dual personality can
be attributed to each of the four surrounding female figures. Walters
(ibid.) suggests that the figures are semi-detached for reasons of allegorical
narrative. They are seasons, but they are also goddesses representing the
cycle of life, death, and resurrection. All of the animals, Dionysiac in fonn, SYMBOLS OF POWER AND NATURE: ORPHEUS MOSAICS
2. Cirencester (2)
4. littJecote Park
5. Newton 51. Loe
12. Pit Meads
Figure 8.6. Distribution map of Orpheus mosaics.
were also intended to convey the same theme.
In order to pursue these ideas further, the mosaics have to be placed
within the social/ architectural context of their construction and use.
THE HIsTORICAL CONTEXT —-
The early fourth century saw a ‘flowering’ of villas in Britain, particularly in
the south-west. A number of factors were responsible for this move to the 114
Table 8.1. Summary of the evidence for Romano-British Orpheus mosaics.
DEFINITE Description Date
Barton Farm Concentric circles coins c.293
Withington Concentric circles c.325-50
Woodchester Concentric circles c.300-25
Newton St. Loe Concentric circles c.325-50
Littlecote Park Radial (circular) c.360
Radial (circular) c.350
Radial (circular) c.350
Medallion 4th century
Concentric circles 4th century
Square (concentric) c.350-60
Uncertain 4th century
Circular (?) 4th century
Rectangular panels ?
Fragments 4th century
countryside, one of the most important being the problem of inflation in
the third century, and the debasement of the silver coinage from Caracalla
onward. At this time the state preferred to collect taxes in kind, and the
larger landowners were obviously the best equipped to deal with the SYMBOLS OF POWER AND NA1lJRE: ORPHEUS MOSAICS
__ ORPHEUS __
Figure 8.7. Orpheus representations schematically interpreted.
problem of taxation. As a result of this, a rigidly ordered society was being
created, induced by the state’s need to gather as much revenue as possible,
and by the tendency of the territorial magnate to increase his estate by
taking over the land of those who were not rich enough to survive in the
It is clear that, by the fourth century, the villas had become the primary
centres for status display. As Millett (1990: 197) notes, those clients and
others who required services from the powerful would come to them for
an audience, and in all probability this would have been at their rural residences. The incentive for public display in the towns had disappeared by the
early fourth century, and the later Romano-British villas should be seen as
a re-direction of the surplusses of society towards personalised rather than
communal display (ibid.).
Scott (1990: 169) has suggested that general trends in villa design, such as
symmetrical facades, wings, courtyards, and enclosures, are related to the
changing social situation noted above. She points out that these features
first made their appearance in villa design in the second century, at about
the same time as the establishment of a market economy, with the scale
and degree of formalisation of these design elements reaching a peak in
the fourth century. Scott (ibid.: 170) suggests that these architectural
features express a duality of purpose. On the one hand they represent a
sophisticated attempt at entry into the Romanised world of markets and
‘civilisation’, and on the other, they are an attempt to distance the household from an environment thought to be potentially hostile. This hostile
environment was not just the perceived physical threat of barbarians.
There was, Scott (ibid.) suggests, a more insidious threat: vulnerability to 116 S.SCOTI
market forces and therefore poverty, inflation, taxation, and the need to
accept strangers over the threshold.
An analysis of those villas possessing Orpheus mosaics can, I suggest,
contribute a great deal to the argument outlined above. In particular, it will
serve to hlghlight a number of changes that were taking place throughout
the fourth century in terms of the manipulation of architecture within
Concerning the villas and mosaics constructed in the earlier part of the
fourth century, it does seem that the sites fit in with Scott’s argument. The
architecture obviously aims to impress, while also allowing the villa owner
to control access to and within his home. Woodchester (Figure 8.9), for
example, possesses three courtyards, and symmetry was obviously a major
consideration in its design (Clarke 1982). Clarke (ibid.) notes that the site
of the villa rises slightly towards the inner courtyard and room 1, and these
features must have formed a and architectural climax.
Additionally, room 1 was vast; nearly 50 ft. (15m) square, with its floor
entirely covered with the Orpheus pavement. As to the superstructure,
Clarke (ibid.) suggests that a domed roof would be consistent with the concentric design of the pavement and the presence of columns, but the walls
seem insufficiently buttressed, especially to the south. Another possibility is
that the columns supported a gallery, and if so, this would have enabled
the pavement to have been appreciated to best advantage. As yet, however,
there is no evidence for a staircase. Both of the possibilities outlined above
would have enhanced the impact of the Orpheus pavement. An important
point to note is that room 1 most probably had four entrances, one in the
centre of each wall. It would almost certainly have been the first room that
was entered by guests. I suggest, therefore, that it would probably have
been used for reception purposes; a form of reception hall where the
patron could meet his clients, and perhaps also the location for large scale
banquets. On entering the corridor (room 2), the visitor would have been
on a visual axis with room 1, with Orpheus, and possibly a central
fountain,4 in view. Once in the room, the visitor may have been able to
view the other corridors leading away to further wealth and splendour.
Those guests who were on a sufficiently intimate footing with the owner
may have been invited to experience this wealth beyond. The ‘depth’ to
which a guest was able to penetrate the building, and the route that he
took, emphasised the nature of his relationship with the owner.
Ellis (1991: 127) has noted that in late antique houses, statues of gods
and mythological heroes were often employed to glorify the owner. Mosaics 10
[Q  … I Dyn.m” ‘p’o, I’
Likely main entrance
Figure 8.8. Viewing patterns at Woodchester villa.
‘-l 118 S. Scurr
were often used for the same purpose, for example, in North Africa many
home owners commissioned pavements with hunt scenes, in which the
owner himself was portrayed as the hero of the hunt. It is not unlikely that
the owner of the villa at Woodchester chose Orpheus for similar reasons.
Orpheus was able to control nature in its strongest and wildest forms without the use of physical force, and would therefore have been an appropriate choice for a room in which the owner would have conducted business,
and entertained friends and/or strangers, and generally aimed to impress.
The villa owner was associating himself with godly powers. As Ellis (1991:
126) notes, such interpretations might seem farfetched were it not for
Ammianus’s description of the flattering comparisons made to senators in
Rome. Additionally, Diocletian attempted to raise the office of emperor to
a transcendent plane, as did Constantius 2 in fourth century Rome (ibid.:
129). If the emperors could be seen as gods, then it is not unlikely that a
villa owner could associate himself with a mythical hero.
It seems that other less wealthy villa owners in the area were sufficiently
impressed by the Orpheus pavement at Woodchester to commission such
pavements for themselves. Although they may not have possessed such cons.iderable means, they certainly had similar ambitions and aspirations.
Based on the chronology outlined in Table 8.1, it could be suggested that
’emulation’ (Miller 1982) was taking place. The owner of the villa at Woodchester could be seen as a form of ‘local emperor’. In order to improve
their position within the social hierarchy, other villa owners in the area, at
Barton Farm, Withington, and Newton St. Loe, for example, may have
adopted this powerful symbol.
The villa at Littlecote represents, I think, a further development of these
ideas. The room possessing the mosaic is, in fact, separate from the main
villa building. It also lacks heating, and would probably have been cold and
damp given its situation by a river. Walters (1982) has suggested that this
hall would have been used for ceremonial purposes, perhaps the meetings
of an unrecognised sect or fraternity in fourth century Britain. I would
suggest, however, that this hall may instead have been a reception hall.
Such chambers have been identified by Ellis (1988) in Roman and provincial houses, and they consist of large apsidal rooms, preceded by a
vestibule, that can be entered through the main door of the house onto the
street. The location of the halls next to the street would have ensured the
maximum amount of privacy for the family. The aristocrat presumably
appeared in the apse surrounded by his retainers (Ellis 1988: 569). Bek
(1983: 91) has noted that in imperial palace architecture the apse was en-SYMBOL<; OF POWER AND NATURE: ORPHEUS MOSAICS 119
visaged as a backdrop for the emperor. This idea of the apse as a backdrop
for persons or events, rather than sculptures or furniture, may be relevant
in the case of Littlecote. The ‘hall’ possesses both a vestibule and apses,
and was also separate from the main villa building. It may have been here
that the owner met his clients, perhaps appearing in the apse at the far
end. The villa owner kept his public and private life separate, and perhaps
only a privileged few would have been admitted to the main building. The
mosaic itself, with its complex religious images, would have emphasised the
formality of the architecture and the superiority of the villa owner. Those
visitors who lacked the necessary education would have been excluded
from the significance of the design, and their social distance from the villa
owner would have been further emphasised.
It is interesting to note that the villas at Horkstow and Whatley have
similar large halls, and both date to around the mid-fourth century, or
slightly later. Like Littlecote, the Horkstow pavement is complex in its
imagery, and combines Orpheus with a number of other mythological subjects.
These developments in domestic architecture coincide with the developments in religious thinking discussed briefly above. Looking back to the
evidence for the ‘hunter god’, it is possible to suggest that the elite of midfourth century Britain felt it necessary to promote religious ideas in which
various gods and deities were conflated, and in which the emphasis was on
power over nature. This pattern fits in with changes that were occuring
elsewhere in the empire. As Henig (1986: 194) suggests:
we must not ignore the deepening religious response of the fourth
century Roman (and provincial) aristocracies in the fourth century,
pagan as well as Christian. For the emperor Julian, Homer, Vergil
and other Greek and Roman authors were writers of ‘holy writ’.
The pagan resistance was undoubtedly widespread, but its core in the
west was the Roman senate which, after Rome had ceased to be the capital
of the Roman empire, assumed once more in Roman history a conspicuous
role (Bloch 1963: 194). Notable evidence for such a revival in Britain
occurs in the form of the Mildenhall Treasure (Painter 1977) and the Corbridge Lanx (Haverfield 1914), both of which appear to date from around
the reign of Julian (355-363), and whose pagan character can hardly be
disputed. By commissioning pavements with complex pagan images, the
owners of the villas at Littlecote and Horkstow may have been associating
themselves with these religious elites elsewhere in the empire, while also 120 S. SCOTf
asserting their own individuality through their inventive syncretism.
Ellis (1988: 573) proposes that the architectural developments may be
associated with two historical trends: the concentration of wealth and
power in the hands of a few aristocrats, and a change in the form of
personal patronage. According to Sidonius (ibid.: 575), mid-fifth century
Gaul appears to have been dominated by a villa-based aristocracy, with
more autocratic relations between patron and client. The government was
investing more power in local aristocrats, and in 371 landlords were made
legally responsible for the collection of taxes. The only way that the poor
could resist demands from officials and aristocrats was by resorting to
someone with more power, even if they risked losing their liberty in the
process (ibid.: 576). The construction of a hall such as that at Littlecote
would have helped the villa owner to assert and maintain his authority. The
patron could have appeared ‘godlike’ in the apse, surrounded by complex
and impressive decor. The separation of the room from the main villa
would have distanced the owner from the ‘domestic’ scene, stressing the
contrast with those visitors whose main concern would have been domestic
issues and survival in the changing world. The ‘rules’ that governed meetings between patron and client, and the repetition of such ‘rituals’, would
have meant that these social relations were constantly reinforced through
In summary, it has been suggested that the Orpheus pavements of the first
half of the fourth century represent overt statements of power on the part
of the villa owners. The nature of the designs would probably have enabled
visitors or residents of the villa, from all social backgrounds, to comprehend the central figure as a figure possessing power. The Orpheus
pavements are commonly found in reception or dining rooms at this time,
and would almost certainly have been the rooms seen first by guests.
Guests may have obtained glimpses of the rooms beyond, and according to
their degree of intimacy with the owner may have been invited ‘deeper’
within the villa. The divisions between the ‘public’ and ‘private’ areas of the
villas seem to have become more clear cut after the mid fourth century. In
some cases visitors may not have been allowed to penetrate the villa at all,
and were received instead in a separate hall, or audience chamber. The
power of the villa owners may have become more autocratic by this time,
and the architecture would have removed them from the domestic setting SYMBOlS OF POWER AND NATURE: ORPHEUS MOSAICS 121
into an almost religious context. The complexity of the designs would have
emphasised the social differences between patron and client, and these differences would have been reaffirmed through the everyday use of the
One other point to note is that the villa owners themselves may have had
apprehensions regarding their own security, both in this life and the next.
The increasing concern with complex religious ideas may have been an
attempt on their part not only to assert their superiority within the social
order, but also to associate themselves with various deities and to obtain
their protection. The conflation of numerous gods may have increased
their feelings of security in a period when empire wide problems were
making their future seem increasingly uncertain.
1. V.C.H. Somerset I, 1906,317.
2. V.C.H. Somerset I, 1906,312-314.
3. B. Walters – Paper presented at the 25th Symposium on Roman Mosaics, 7th Dec. 1991.
4. Research carried out by the Woodward brothers for a reconstruction of the pavement
suggested water staining around the centre of the pavement, which may have been caused by a
Beecham, K. J. 1886. History of Cirencester and the Roman City of Corinium.
Bek, L. 1983. Questiones Convivales: the Idea of the TricJinium and the
Staging of Convivial Ceremony from Rome to Byzantium. Analecta
Bloch, H. 1963. The Pagan Revival in the West at the End of the Fourth
Century. In A. Momigliano (ed.), The Conflict between Paganism and Christianity in the Fourth Century, 193-218. Oxford.
Boas, Franz 1955. Primitive Art. New York: Dover.
Brendel, O. J. 1979. Prologemena to the Study of Roman Art. New Have: Yale
Buckman, J. and C. H. Newmarch 1850. The Remains of Roman Art in
Clarke, G. N. 1982. The Roman Villa at Woodchester. Britannia 13:197-228.
Clifford, E. M. 1938. Roman Altars in Gloucestershire. Transactions of the
B.G.A. Society 60:297-307.
Cookson, N. 1984. Romano-British Mosaics. Oxford: British Archaeological
Reports (British Series 135). 122 S.SCOTI
Coward. R. and]. Ellis 1977. Language and Materialism. R.K.P.
Ellis. Simon P. 1988. The End of the Roman House. American Journal of
Ellis. Simon P. 1991. Power. Architecture. and Decor: How the Late Roman
Aristocrat Appeared to His Guests. In E. Gazda (ed.). Roman Art in the
Private Sphere. 117-134. Ann Arbor: University of Michigan Press.
Frere. Sheppard S. 1982. The Bignor Villa. Britannia 13: 135-195.
Goodburn. Roger 1979. Chedworth: The Roman Villa. London.
Haverfield. R. 1914. Roman Silver in Northumberland. Journal of the Roman
Henig. Martin 1986a. Ita intellexit numine inductus tuo: some personal
interpretations of deity in Roman religion. In Martin Henig and A. King
(eds). Pagan Gods and Shrines of the Roman Empire. 159-169. Oxford
University Committee for Archaeology Monograph 8.
Henig. Martin 1986b. Late Roman Mosaics in Britain: Myth and Meaning.
Hinks. R. 1933. Catalogue of Greek, Etruscan and Roman Paintings in the British
Hoare. R. C. 1819. History of Ancient Wiltshire. Part 2. Roman Aera. London.
Hodder. Ian R. 1986. Reading the Past. Cambridge: Cambridge University
Ingram. J. 1849. Notices of the Mosaic Pavement discovered at Thruxton.
Hants. in 1823. Proceedings of the Royal Archaeological Institute. Vol. of the
Wilts. and Salisbury meeting. 241-245.
Jesnick. I. 1989. Animals in the Orpheus Mosaics. Mosaic 16:9-13.
Ling. R. 1982. Mosaics and Murals. Mosaic 7:26-28.
Lysons. S. 1817. Roman Antiquities at Woodchester.
McWhirr. Alan 1986. Houses in Roman Cirencester. Cirencester Excavations 3.
Merrifield. R. 1986. The London Hunter-God. In Martin Henig and
Anthony King (eds.) Pagan Gods and Shrines of the Roman Empire. Oxford
University Committee for Archaeology Monograph 8. 85-92.
Miller. Daniel 1982. Structures and Strategies: an Aspect of the Relationship
between Social Hierarchy and Cultural Change. In Ian Hodder (ed.)
Symbolic and Structural Archaeology. 89-98. Cambridge: Cambridge
Millett. Martin 1990. The Romaniz.ation of Britain. Cambridge: Cambridge
Morphy. Howard (ed.) 1989. Animals into Art. London: Unwin Hyman.
Neal. D. S. 1981. Roman Mosaics in Britain. Gloucester.
Nichols. W. L. 1938. The Roman Villa at Newton St. Loe.
Painter. K. S. 1977. The Mildenhall Treasure: Roman Silver from East Anglia.
London: British Museum Publications. SYMBOU; OF POWER AND NATURE: ORPHEUS MOSAICS 123
Preston Blier, S. 1987. The Anatomy oj Architecture. Cambridge: Cambridge
Rawes, B. 1977. A Romano-British Site on the Portway. Glevensis 11.
Reece, Richard 1980. Town and Country: the End of Roman Britain. World
Archaeolog;y 12 (1):77-92.
Rivet, A. L. F. (ed.) 1969. The Roman Villa in Britain. London: Routledge.
Salway, Peter 1984. Roman Britain. Oxford: Oxford University Press.
Scott, Eleanor 1990. Romano-British Villas and the Social Construction of
Space. In Ross Samson (ed.), The Social Archaeolog;y oj Houses, 149-172.
Edinburgh: Edinburgh University Press.
Scott, Sarah 1989. Symbols of Power and Nature: a Contextual Approach to
the Orpheus Mosaics of Fourth Century Roman Britain. Unpublished
BSc. dissertation, University of Leicester.
Scott, Sarah 1991. An Outline of a New Approach for the Interpretation of
Romano-British Villa Mosaics, and Some Comments on the Possible
Significance of the Orpheus Mosaics of Fourth Century Roman Britain.
Journal oj Theoretical Archaeolog;y 2:29-35.
Shanks, Michael and Christopher Tilley 1987. Social Theory and Archaeolog;y.
Cambridge: Polity Press.
Smith, D.]. 1969. The Mosaic Pavements. In A. L. F. Rivet (ed.), The Roman
Villa in Britain, 71-125. London: Routledge.
Stead, Ian M. 1976. Excavations at Winterton Roman Villa and other Roman
Sites in N. Lincolnshire. Department of the Environment (Report No.9).
Stevens, C. E. 1933. Sidonius Apolinaris and His Age. Oxford.
Thebert, Y. 1987. Private Life and domestic Architecture in Roman Africa.
In P. Veyne (ed.), A History oj Private Life, 1, From Pagan Rome to
Todd, Malcolm 1978. Villas and Romano-British Society. In M. Todd (ed.),
Studies in the Romano-British Villa, 197-208. Leicester: Leicester University Press.
Toynbee, Jocelyn M. C. 1962. Art In Roman Britain.
Toynbee,Jocelyn M. C. 1964. Art in Britain under the Romans.
Walters, B. 1981. Littlecote. Current Archaeolog;y 80:264-268.
Walters, B. 1982. Fourth-Century ‘Orphic’ Halls in Britain. Mosaic 7:23-26.
Wedlake, W.]. 1982. The Excavation oj the Shrine oj Apollo at Nettleton,
Wiltshire, 1956-1971. London: Society of Antiquaries.
Wheeler, R. E. Mortimer and Tessa V. Wheeler 1932. Report on the Excavations oj the Prehistoric, Roman and Post Roman Site in Lydney Park,
Gloucestershire. Oxford: Society of Antiquaries of London.