The Association for the Study and Preservation of Roman Mosaics

About ASPROM News & Events Publications Mosaics in Britain Resources Join ASPROM

Mosaic in Lullingstone Villa (photo by David S. Neal).
Mosaic in Lullingstone Villa (photo by David S. Neal).

Art, religion and letters in a fourth-century villa:
the Lullingstone Villa mosaic

Martin Henig

The figured mosaic at Lullingstone Villa, Kent is one of the best-known mosaics of Roman Britain. It features two mythological scenes juxtaposed: Bellerophon slaying the Chimaera (in an anteroom),[1] and Europa on the bull which is, in fact, Jupiter in disguise (in the triclinium apse).[2] The floor is generally dated to around AD 330, but as (in the view of this writer) there is a general similarity to the work of the Durnovarian school (note the fat, slug-like dolphins around Bellerophon and the lithe and elongated forms of Pegasus and the bull), this might suggest that it was laid in the middle of the fourth century and perhaps as late as c. 360 by artists from the south-west.[3]

The verse inscription

Lullingstone Villa: scene of Europa and the bull, with inscription above (photo by David S. Neal).
Lullingstone Villa: scene of Europa and the bull, with inscription above (photo by David S. Neal).

The Europa scene is accompanied by an elegiac couplet which recalls Ovid’s structure, metre and syntax.[4] The story of Europa and the Bull is to be found in the Metamorphoses (II, 11, 835–75) and although, on the Lullingstone mosaic, Europa looks far more self-assured than she is in the poem where, half-terrified, she clutches the animal’s horn and his back, the outline technique does allow a correspondence with the bull’s colour ‘white as the untrodden as yet unmelted by the damp south wind’. The iconography, which includes two cupids, one holding the tail of the bull while the other leads from the front, is not unique to Lullingstone and is also to be found on a somewhat later mosaic (c. AD 400) from Djemila (Cuicul) in Algeria.[5]

The verse makes direct allusion to the famous episode at the beginning of the Aeneid where Juno visits the west wind (Aeolus) to ask him to brew up a storm in order to drown her enemy (I, 50). The Aeneid was well-known in Roman Britain and, indeed, the aftermath of the storm which led to the Romans being washed up on the shores of Dido’s realm at Carthage is the theme of the Low Ham pavement. This mosaic is unique in the Roman Empire and was very probably copied from the owner’s illustrated copy of the work.[6]

It does not require much effort of imagination to see the Lullingstone couplet too as a product of the owner’s education and taste. In his schooldays he would certainly have studied Vergil and Ovid in depth, and learned the art of versification. A villa’s triclinium, a place of relaxation (otium) where a traditional literacy culture was allowed to flourish, was the proper place to impress one’s guests with such scholarly accomplishments. Here these included a mildly erotic joke based on a contrast between the events in two poems and even, perhaps (if the Ovidian reference cited above is taken into account), the drawing of a contrast between the fierce action of the west wind and the gentle balm of the south wind; a contrast between the destruction wreaked by the one and the calming effects of the other.

I have always believed that such mythological mosaics, for the most part, proclaimed the pagan beliefs of their owners.[7] The discovery that follows, only partly mine, while it strengthens the claim that such displays of figural art in Late Antiquity were likely to have been highly charged with meaning, provides a salutary warning against imposing too rigid a division between pagan and Christian tastes.

Several scholars, notably David Howlett,[8] Kenneth Painter and Professor Charles Thomas, have become interested in recent years in the sophisticated latinity of insular writing, most of it, of course, post-Roman in date but including the works of Pelagius and Patrick, who may be regarded as Roman Britons. Like other Christian authors they believed (as indeed did Jews and Greeks) that creation followed mathematical rules and this led them to use various numerical sequences and relationships in their compositions. Sacred numbers such as 5 and 7 were clearly of account here. The source of this note lies in a letter I received from Professor Thomas last year, detailing the existence of a cryptographic element in a few of the Early Christian (fifth to eighth century) gravestones of western Britain and asking me whether I had come across anything similar in Roman inscriptions.

The problem here is that there are very few surviving fourth-century inscriptions from Roman Britain. Only the verses on the Lullingstone mosaic and on that from Frampton (see below) with their elusive imagery cry out for such interpretation. I had a quarter of an hour before lunch on the day upon which I received this request and fortunately decided to begin with Lullingstone

I counted five letters to the end of the first word INVIDA, a term which was used as a prophylactic against the Evil Eye (Invidia). I then proceeded with several letters at a time:


to find a name, AVITUS. It might be observed that the name taurus is close to being an anagram of Av(i)tus and one wonders whether the playful bull was not seen as representing the host.

What of the rest of the line? Professor Thomas believes that the reading ends with aeolias. If one continues:


there is a V and an O followed immediately by an S, though it might be significant that if one returns to the beginning of the first line you get an S in the regular sequence, having passed the ominous word invida:


In any case one can find the pronoun VOS, meaning ‘(all of) You’.

The context in which such a trick might be understood was given to me by Professor Thomas, who rightly suggested that I should not have left it with finding a personal name but continued to examine the text for its hidden meaning; there was more to the inscription than this.

Both lines begin with an I and end with an S. In the Middle Ages the letters IHS are employed as a sacred monogram; might IS stand for IesuS here?

The second line, in fact, allows one to count in 7s until the last gap (a count of five):


to find the five-letter name of Jesus in full. Professor Thomas further suggested to me that the word Deus was very probably present as well, here using double 7s as intervals in:


Commentary and discussion

The name Avitus is fairly common in Late Antiquity and is most plausibly that of the villa-owner: was he perhaps a forebear of that prominent Gallic family, connected with Sidonius Apollinaris’ in the mid-fifth century and including Eparchius Avitus, for a short time an emperor (455–6) before being deposed and consecrated as bishop of Placentia?[9] The only putative villa owner called Avitus previously attested in Roman Britain is Ausonius Avitus whose name appears in a niello inscription upon a bronze handle from Broomhill. East Winterslow, Wlitshire, whose decoration looks fourth-century in date.[10] The reading, if accepted, gives a second named villa-owner upon a mosaic after that of Quintus Natalius Natalinus on the floor from Thruxton, Hampshire, now in the British Museum.[11]

Of wider interest and significance is the discovery of a Christian meaning hidden in the Lullingstone pavement. This does not come as a total suprise for, after all, the next generation (c. 380) saw a house-church at the villa, attested by frescoes in an upper suite of rooms, incorporating chi-rhos and a procession of Orantes.[12]

It is especially interesting to see how two religious and cultural traditions, pagan and Christian, were fused together in a way that extreme pagans (one thinks of Julian) and the most austere Christians would certainly not have liked.[13] Avitus of Lullingstone seems to tell us that he has achieved salvation with the god (Deus), Iupiter/Christ, while in contrast You (vos) face the justice of destruction unless you follow him as though over the sea!

The standard of intellectual dexterity displayed in the literary choice of source and subject matter, cryptography and the meaning is very high; it is by no means perfect, but would it be reasonable to expect more of an educated amateur? Avitus was showing off his skills to a closed, inner group of friends and dining companions who met in his triclinium, while his servants and casual acquaintances remained in ignorance.

The other figural scene in the adjoining room, Bellerophon slaying the Chimaera, is the subject of an important paper by Anthony Beeson in the last issue of Mosaic. He points out that Pegasus was the child of Neptune and that Bellerophon could represent the power which controls the seasons, the sun, as is confirmed by an inscription on the Bellerophon mosaic in the so-called Palace of Theodoric at Ravenna. The Bellerophon theme has sometimes been given a Christian gloss; though the evidence for this with regard to Lullingstone has previously been weak, depending on later Christian wall-paintings. Now it seems that the scene is juxtaposed with a Christian inscription in the same way that the Bellerophon episode is connected with a bust of Christ and a chi-rho at Hinton St Mary, Dorset, and with a simple chi-rho at Frampton in the same county. If in three out of four occurrences of the scene in Roman Britain there is a direct Christian connection, the case for such an interpretation being commonplace in the province amongst educated Christians becomes much stronger, if it is not overwhelming. Even the Bellerophon mosaic at Croughton, Northamptonshire, which does not appear to be joined to any other scene, has a line of stars, representing the Heavens, appearing above horse and rider. It might be noted that the panel with the Lullingstone Bellerophon is framed by the Seasons who are normally controlled by the Sungod though the Ravenna inscription (cited by Beeson), ‘I am he through whom autumn, winter and summer are in turn restored’ would suit Christ. Immediately surrounding the scene are marine motifs, rather slug-like dolphins and bivalve seashells, matching the marine setting of Europa’s abduction.[14]

There may be a previously unnoticed link with the main Frampton floor here. The Frampton mosaics have always seemed a problem, because of the overwhelmingly pagan nature of the rest of the imagery, with the gods Bacchus and Neptune each invoked on two mosaics and a range of mythological themes, some of which could have been derived from Ovid.[15] There is of course also an intriguing verse inscription (in catalectic anapaestic dimeters) which, alas, has not yet yielded its inner secrets to me — if it has any —, neither owner’s nor divine name. However. even on a surface reading we find the ‘head of Neptune . . . whose dark-blue figure is flanked by two dolphins’ and who rules ‘the domain stirred by the winds’, a reference to the Ocean which we must all cross and perhaps to the contrasting west and south winds. There is another (fragmentary) verse naming Cupid, perhaps meaning the young Bacchus, though we should also recall the two cupids shown with the Lullingstone bull. Here is the active force of salvation. While the imagery, both literary and artistic, is a guarantee that the ambience is one in which the old Graeco-Roman cults were taken seriously, is it not possible that far from ‘paganising Christ’ they were interpreting the old faith to make it acceptable to insular Christians? The figure of Bellerophon implies the power of Christ and perhaps even the ability of the heroic Christian to smite evil (the Chimaera) through his own efforts. If so it is Pelagian in its implications a generation before Pelagius.[16]

Such a faith could co-exist with paganism. At Lullingstone we find not just the deeds of a god and a hero on the mosaic, but also evidence of offerings to the house-spirits (represented by two second-century marble busts) in the cellar. At Frampton Christ–Bacchus was the saviour god, his power exemplified by pagan, literary culture. This is an example of the mixed cultural beliefs of cultivated aristocrats like Ausonius, tending towards syncretism; not of the conventional orthodox theology of austere churchmen such as Ambrose. It fitted into a highly sympathetic, literary, aristocratic culture in which the old gods were still regarded alongside Christianity, where, in the words of Quintus Aurelius Symmachus, ‘not by one path alone can man comprehend so great a mystery’ (Relatio 3,9 ).[17]

The primary interests of this Association are very properly with mosaics. Amongst inscriptions in this medium, the floor of the cella of the Temple at Lydney Park, Gloucestershire, laid at about the same time as the Lullingstone pavements and in this case dedicated by the praepositus religionis of Mars Nodens, Titus Flavius Senilis, is evidence of the continued vitality of paganism. Presumably the geometric floors in the guest house here were under his patronage or at least that of his co-religionists.[18] The owner of the Roman villa at Thruxton in Hampshire, had his name, Quintus Natalius Natalinus, inscribed on his dining-room floor, probably dedicated (ex voto) to Bacchus who is depicted in mosaic.[19] Other non-inscribed, figural mosaics appear to be pagan or at least religiously neutral. The mosaic floor of the triconch hall at Littlecote, Wiltshire appears to me (as to the excavator) to expound a sacred drama which reconciles Apollo to Bacchus through Orpheus. The complex of mythological and religious scenes at Brading, Isle of Wight, refer to the mysteries of Demeter, Cybele, Bacchus and Apollo, to Orpheus and either to Abraxas–Iao or to Hermes Trismegistos as well as to other Late Antique eschatological ideas such as theurgy. The Mysteries Mosaic at Trier shows that such esoteric cults were not uniquely insular; the Thetford Treasure, jewellery and plate belonging to a collegium of Faunus, reminds us that such evidence is not confined to mosaics.[20]

The Lullingstone mosaic presents us with themes as difficult and elusive for our full comprehension today, as those encountered at Frampton and Hinton. They suggest that the owner of the Lullingstone villa in the mid-century was a Christian twenty years or so before the house-church was in use; he was also highly educated in the ancient traditions of grammar and rhetoric, and had by no means turned away from the ancient gods. The mosaic implies that we need to view insular (Romano-British) Christianity as both special and different. It is a further demonstration of culture and education alongside the surviving writings of Patrick and other insular Latin authors, which do show some knowledge of the popular classics. It confirms the suggestion that the great books of Rome (and perhaps even of Greece) continued to be read in Britain and Ireland.[21] It allows us to take ‘Dark Age’ insular traditions of composition at least as far back as the middle of the fourth century, long before Jerome’s Vulgate and thus to show a continuity of culture from Roman times to the Middle Ages, long beyond the period in which mosaics were laid in Britain. The Europa mosaic is thus to be seen not simply as an interesting pavement but as a work of crucial importance in the long — and continuous — story of culture in these lands.

I am most grateful to Dr David Howlett and to Professor Charles Thomas and to Mr Kenneth Painter for stimulating any interest in a fascinating area of Late Antique culture. Perhaps it is time for more archaeologists to stop digging and start counting!
This paper is limited to the Lullingstone mosaic and does not aim to anticipate their fuller studies in the subject which take us beyond the world of mosaics.


1J. M. C. Toynbee, Art in Britain under the Romans (Oxford 1964), 264–5, pl. lx b; G. W. Meates, The Roman Villa at Lullingstone, Kent, I: The Site (Maidstone 1979), 78–80, and col. pl. General discussion of the mosaic, pp. 82–3.
2J. M. C. Toynbee, Art in Britain under the Romans (Oxford 1964), 263–4, pl. lx b; G. W. Meates, The Roman villa at Lullingstone Kent, I: The Site (Maidstone 1979), 75–8, colour frontispiece.
3See M. Henig, ‘Late Roman mosaics in Britain: myth and meaning’, Mosaic 13 (1986) 13–20, esp. pp. 18–19; M. Henig, The Art of Roman Britain (London 1995), 125, see ill. 77.
4A. A. Barrett, ‘Knowledge of the literary classics in Roman Britain’, Britannia 9 (1978), 307–13, esp. p. 311. For the definitive publication of the verse see S. S. Frere and R. S. O. Tomlin in R. G. Collingwood & R. P. Wright, The Roman Inscriptions of Britain II, fascicule 4 (Haverfield Bequest, Stroud 1992), (henceforward RIB II 4), 86 no. 2448.6.
5M. Robertson in Lexicon Iconographicum Mythologiae Classicae IV (Zurich & Munich 1988), 85, pl. 44 no. 162 (Lullingstone) and no. 164 (Djemila).
6J. M. C. Toynbee, Art in Britain Under the Romans (Oxford 1964), 241–6, pl. lviii; M. Henig, The Art of Roman Britain (London 1995), 126, col. pls ix & x. The widespread knowledge of Vergil in Late Roman Britain is emphasised by the fact that Carausius could even use the initial letters of a catch-phrase in Eclogue IV as a mintmark (RSR = Redeunt Saturnia Regna). See Guy de la Bedoyere, Current Archaeology XIII, no. 9 (1997) 358.
7M. Henig, ‘Late Roman mosaics in Britain’, and idem, ‘ita intellexit numine inductus tuo: some personal interpretations of deity in Roman religion’, pp. 159–69 in M. Henig & A. King, Pagan Gods and Shrines of the Roman Empire, Oxford Committee for Archaeology Monograph No. 8 (Oxford 1986); idem, The Art of Roman Britain (London 1995), 159–60.
8D. R. Howlett, The Book of Letters of Saint Patrick the Bishop (Dublin 1994); idem, The Celtic Latin Tradition of Biblical Style (Dublin 1995); see pp. 56–65 for Pelagius.
9A. H. M Jones, J. R. Martindale & J. Morris, The Prosopography of the Later Roman Empire I, AD 260–395 (Cambridge 1971), 127 for a mid fourth-century orator of which little is known; J. R Martindale, The Prosopography of the Later Roman Empire II, AD 395–527 (Cambridge 1980), 194–5 for a fifth-century Gallic senator, relation and friend of Sidonus Apollinaris, and 196–7 for Eparchius Avitus, emperor AD 455–6.
10RIB II, fasc. 3 (Haverfield Bequest, Stroud 1991), 78 no. 2433.3
11RIB II, fasc. 4, 90–1 no. 2448.9; M. Henig and G. Soffe, ‘The Thruxton Roman Villa and its mosaic pavement’, Journal of the British Archaeological Association 146 (1993) 1–28.
12J. Liversidge in G. W. Meates, The Roman Villa at Lullingstone Kent, II: The Wall Paintings and Finds (Maidstone 1987), 11–49.
13G. Fowden, Empire to Commonwealth: Consequences of Monotheism in Late Antiquity (Princeton NJ 1993), ch. 4, esp. pp. 86–92 on Constantine’s mission to Christianise the Empire; by contrast see pp. 86–92 on Constantine’s mission to Christianise the Empire; by contrast see P. Athanassiadi, Julian: An Intellectual Biography (London, 2nd edn, 1992), ch. V, pp. 161–92 for the pagan challenge to the policy of imposed monotheism.
14A. Beeson, ‘Pegasus the Wonder Horse and his portrayal on Romano-British mosaics’, Mosaic 23 (1996), 18–23: C. Lochin, ‘Pegasos’, in Lexicon Iconographicum Mythologiae Classicae VII (Zurich & Munich 1994), 225 no. 171 (also Lullingstone), no. 170 (Frampton); see p. 219 no. 64 for a Christian sarcophagus in Rome and p. 230 for Bellerophon as a ‘Christian’ theme. J. Huskinson takes a fairly sceptical line with regard to ‘Bellerophon Christianus’ in ‘Some pagan mythological figures and their significance in Early Christian art’, Papers of the British School at Rome XLI I (1974), 68–97, and especially pp. 75–8 for the British evidence.
15M. Henig, ‘James Engleheart’s drawing of a mosaic at Frampton’, Proceedings of the Dorset Natural History and Archaeological Society CVI (1984), 143–6.
16For the inscription, RIB II 4, 88–9, no. 2448.8. For the theory that the Bellerophon myth was in accord with the development of a ‘Pelagian’ theology see M. Henig, The Art of Roman Britain (London 1995), 156.
17See E. W. Black, ‘Christian and Pagan hopes of salvation in Romano-British Mosaics’, pp. 147–58 in Henig & King, Pagan Gods and Shrines of the Roman Empire, esp. pp. 148–50. His interpretation now seems closer to the truth than my own in ‘ita intellexit’, ibid., pp. 163–4. The Christian interpretation is central but suffused with paganism.
18R. E. M. Wheeler and T. V. Wheeler, Report on the excavation of the prehistoric, Roman and post-Roman site in Lydney Park, Gloucestershire, Rep.Res.Comm.Soc. of Ants, London (Oxford 1932), 65–7; RIB II, 4, 84 no. 2448.3.
19See n. 11 above.
20See M. Henig, The Art of Roman Britain (London (1995), 153–4, 163, 168; R. Ling, ‘The iconography of the Brading mosaics’, Mosaic 18 (1991) 14–21 exemplifies that author’s over-sceptical approach: I see no reason to change any view on the basis of this paper, while the Lullingstone discoveries make me even more confident that the educated elite of Roman Britain selected and in many instances were the effective designers of the mosaics which expressed their strongly-held religious and cultural beliefs.
21For insular literature see n. 8; also K. R. Dark, Civitas to Kingdom: British Political Continuity 300–800 (Leicester 1994), 184–91. One wonders whether the great medieval Irish epic The Tain is rather like the Iliad because it was compiled and written by monks who had read the Iliad.

© Martin Henig 1997. This article was originally published in Mosaic 24 (1997) 4–7.
This page is maintained by Ruth Westgate.