The Association for the Study and Preservation of Roman Mosaics

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

Extracts from Newsletter no. 43

October 2003

2003 has been a busy year for those interested in Romano-British mosaics. It is still hoped that Volume II of the corpus of Romano-British mosaics will be published by the end of the year. There have been remarkable numbers of new mosaics found his summer, several having close affinities to known examples and helping in our understanding of the craft’s organisation. The exciting mosaic from Badminton Park is exceptionally large and well preserved, and its unusual design and debased motifs perhaps hint at a very late Roman date. There has also been much activity at sites where cover-buildings protect fine mosaics. Brading is to have its ambitious new structure, and at Fishbourne Palace they are about to launch a fundraising campaign for a Heritage Lottery Fund-supported project to refurbish the North Wing’s cover building.

ASPROM’s Colchester Symposium, 14–15 June

ASPROM members at Colchester, June 2003
Colchester: ASPROM members at the site of the Christian Church

Our Events Secretary, Mary Winsch, writes: ‘41 members booked to come to Colchester, 37 attended. The weekend was capably organised by Dr Paul Sealey, Assistant Keeper of Archaeology. The weather was warm and sunny, the venue welcoming. The Castle, which is the largest Norman keep in England, and one of the earliest, houses the displays on the early history of Essex, from the prehistoric up to the Civil War. The Castle was built on the podium of the Roman temple to Claudius and the temple precinct walls provided the base for the inner bailey rampart. Thanks to Charles Gray the Castle and its grounds were saved from redevelopment and now provide a large and busy green space in the heart of a very pleasant town.

The weekend began with a tour of the Castle, around the outside, into the Roman vaults and up onto the roof. The Normans clearly recycled the stone and brick from some major Roman structures. As there is a layer of pilae bricks half way up the walls, the town baths were the first to go. The theatre probably provided more building material as well as the base for the early chapel to St Helena. The parish church has an Anglo-Saxon tower full of Roman brick, as does St Botolph’s Abbey. Colchester is fortunate in retaining much of its town wall, although there have been some breaches in them, including recent ones. The Balkerne Gate still survives in a modest form.

Colchester has some good mosaics although, as Dr David Neal pointed out, none of them are first century, which is rather odd since there are two famous early military tombstones. The mosaics are on display, some on the walls but one laid flat beneath a balcony which gave a good view of its sea creatures and the birds in the border scrolls. George Monger described his efforts to back one of the mosaics onto a modern aluminium framework. The original backing was concrete that needed a jackhammer to remove it. Paul Sealey described the history of Roman Colchester, its destruction by Boudicca, the rebuilding and then the shrinkage to a line of buildings running along the High Street. He believed that the temple precinct became a late Roman citadel. Outside the walls of the town there is an early Christian church and an extensive cemetery. Philip Crummy talked about the modular construction used by the Romans and how it was possible, using Vitruvius as a guide, to suggest the columnar layout of the Temple and of the rural temple and precinct at Gosbecks. Our thanks go to the four speakers. The Saturday evening buffet was provided by the Really Good Food Company and was enjoyed by all. It was an opportunity to look in detail at some of the wonderful small finds from the town and take photographs of the mosaics.’

What not to do with a mosaic

When David Johnston edited Mosaic, he included a series of short articles entitled ‘What not to do with a mosaic’. What was done to conserve one of the mosaics from Colchester could be considered in this category. Although this is not so much a criticism of the conservator, for he was attempting an untried technique, the result is a lesson in how not to re-back a mosaic.

George Monger spoke at the Colchester symposium about his efforts to conserve the mosaic from North Hill, Colchester. It is part of one of the second-century nine-panel designs so typical of the town. The problem has been one faced by several museums where they have mosaics lifted in Victorian times and set on a heavy concrete matrix. In this case dangerous cracks had appeared, probably caused by the corrosion of the reinforcing iron rods, and, as the piece was mounted on a wall and weighed roughly two tons, remedial action was advisable. At huge expense, the mosaic was removed and transported to George Monger’s workshop in Suffolk; it was to be away from the museum for two years.

Before the recent work, the piece was framed in oak and had a wooden backing to the thick layer of concrete with aggregate and the iron rods; on this was a thinner layer of dense concrete into which the tesserae were set. The wood and the outer layer of concrete were removed relatively easily, leaving irregularly shaped fragments with serrated edges. At this stage the fragments were ‘faced up’—muslin was stuck to the surface of the tesserae with water-soluble glue (PVA). The dense concrete proved more difficult to remove and in the end a pneumatic hammer/drill was used; by this frighteningly vigorous means, most of it was removed, although a few stubborn little islands of concrete were left. This is where an interesting decision was made. Instead of putting the fragments together and dividing it into manageable sections along the straight lines of the design (such as the edge of the bands of guilloche) the choice was made to back the irregular fragments individually! This was probably done because he was following the techniques used by the British Museum in re-backing the Hemsworth mosaic (see The Conservator No. 24 (2000) 61–8)—but the difference was that the Hemsworth fragments did not have to be precisely fitted together afterwards. An adhesive (paraloid paste) was applied to the back of the tesserae, and wooden shuttering and a ‘lid’ were made—not easy considering the awkward shapes with serrated edges. Thus encased, it was filled with epoxy-resin foam. Once the shuttering was removed, the fragments resembled pieces of a jigsaw puzzle waiting to be re-assembled. Almost inevitably, stretching—always a problem when relaying mosaics—and the difficulties with making shuttering for such odd shapes, meant that the re-backed pieces would not fit together. The best fit was made and they were stuck to the lightweight base of aluminium hexlite (a honeycomb arrangement) with Araldite 2015. The gaps were filled with a mixture of adhesive and cement dust collected during the removal of the concrete backing, and the whole framed in aluminium.

Now that the mosaic has been returned to the wall in the Museum at Colchester Castle, the difficulties encountered during the work are only too plain to see. Wide gaps of grey cement grouting show where the fragments joined (or, rather, did not join) and rows of tesserae no longer follow through, with the result that there are steps. In fact, the piece gives a false impression of the Roman workmanship, which now appears careless. While every effort was made to use the best (and extremely expensive) materials, an error in technique has produced a very disappointing result. Hopefully we will learn from this unfortunate episode.

Brading Villa

On Sunday 25 May our President, Rosamond Hanworth, and I went to Newport, Isle of Wight, to take part in a day of lectures to support the fundraising for the new cover-building to protect the extraordinary mosaics at Brading. When we arrived we noticed huge numbers of cars outside the venue and assumed that the hall would be seething with people, but we were soon informed that the swimming pool and other leisure facilities shared the site. The main draw was Phil Harding from Channel Four’s Time Team who gave us some amusing anecdotes about the programme, although he didn’t mention the Brading mosaics—that was left to the ASPROM contingent. The architect explained the concepts behind his design for the new cover-building, which seem very laudable. A great deal of thought has been put into restricting sunlight, and maintaining a constant temperature, drainage, circulation of air (and people!) and providing ancillary services—shop, office, educational facilities and lecture area (perhaps for a future ASPROM meeting?). The roof is particularly ‘eco-friendly’, planted up so that it will certainly blend into the landscape better than the present corrugated iron version, which is at risk of being blown across the island during the next gales. The raising of funds is going well, and it has recently been announced that they are to receive £2.13 million from the Heritage Lottery Fund. It means that work can begin soon.

Bignor Villa

Cover buildings for mosaics are always a problem, and those at Bignor, West Sussex, are no exception. Some of the problems of the superb geometric pavement in Room 6 (see Newsletter no. 35, September 1999) have been addressed—notably the drain running outside the outer wall of the room. It was disappointing to see that since that survey was made, serious structural problems have developed and the roof is currently shored up. Expensive maintenance is difficult especially when the site is in private hands without the resources, for instance, of the National Trust, etc. The land and its villa have been in the possession of the Tupper family since its discovery in 1811. I came across an interesting account of an attempt in 1898 by Richard Tupper, the grandson of the owner at the time of excavation, to sell the mosaics to the British Museum (Proceedings of the Society of Antiquaries of London, 2nd series 18 (1900) 26). The attitude of the time was summed up by the correspondent: ‘Sentimental reasons might be urged for keeping the pavements in situ, but in order to ensure their preservation I feel certain that no better plan could be adopted than their careful removal to the National Collection.’ The sum demanded for the purchase of ‘the Roman pavements, architectural details and other Roman work’ was £1,500; and if they were to be bought, but were to remain at Bignor, an additional £50 per acre for the land. The sum was thought too high, especially when the extra expense of lifting and transportation was considered, and not even a reduced and final offer of £1000 was acceptable. Perhaps this was just as well—the future will tell! Richard Tupper was worried about deterioration such as the lifting of tesserae as the result of frost, and he waged a determined campaign against destructive mice. Nevertheless the splendid mosaics—among the very best in Britain—are still there and a visit to this magnificent site is highly recommended.

Mosaics in the Corinium Museum

The museum has received a substantial grant from the National Lottery Fund to undertake refurbishments. This involves the re-backing of at least two mosaics so that they can be mounted on the walls. Our Honorary Editor, Charles Browne, has told me of his visit to the workshops of Cliveden Conservation at Kilmersdon near Bath, where he saw the work on mosaics from Kingscote and Cirencester. The results appear very satisfactory. They are to be waxed to improve the colour, a process that is reversible.

New mosaic finds

About a dozen new mosaics have been found this summer. It is only possible here to give a brief notice and a fuller description will appear in Mosaic next year. All have been drawn in situ by David Neal (except that at Hawkesbury, which I drew) for the corpus of Romano-British mosaics. Bradford-on-Avon’s will appear in Volume II and the rest in Volume IV.

Bradford-on-Avon, Wiltshire

Bradford-on-Avon: A fifth-century baptistery was built on and through a mosaic in the main room of the villa.

The part of this well-preserved mosaic found last year was re-exposed in July and the previous trench extended to include the rest of the room and the porticus. During the removal of the rubble covering the southern part of the pavement, it became clear that a post-Roman building had been constructed directly over the mosaic after the villa had been demolished. The late structure not only destroyed the centre of the mosaic but also restricted the area that could be excavated. The circles seen at the edge were part of an intermediate panel (as predicted) and the new panel is very similar to that at Brislington, near Bristol, having a scheme of cushions. This is not surprising for, as noted in a previous newsletter, the large cantharus is identical to the one on the Brislington pavement. The mosaic also has stylistic links to some of those from nearby Keynsham and Box, and the flower is identical to one from Bath. This brings the total of mosaics from in and close to the Avon valley to five. One very unusual feature of the newly-discovered panel is the inclusion of guilloche outlined in white instead of the normal dark blue-grey. Only two other examples precisely like it are known—from Room 10 at Woodchester and a fragment from Pitmeads, Wiltshire, already recognised as being by the same hand (see Mosaic 25 (1998) 10, fig. 1). This and other features in the mosaic from Woodchester suggest that all three examples including this unusual technique are attributable to the South-western Group.

Excavations were also continued on the other villa-building close by. Its plan as revealed by a geophysical survey is identical to the first—even the apse of the central room—but unfortunately the floors, if they had survived, may have been removed during the levelling of the playing field under which the remains lie, remarkably close to the surface. It seems unlikely that this second structure ever had mosaics and seems to have had an agricultural function. There is every indication that the site continued to be occupied well into the fifth century.

New book on Bradford-on-Avon villa: Mark Corney, The Roman Villa at Bradford on Avon: The Investigations of 2002. This excellent booklet (24pp) by the director of the dig gives a background to previous discoveries in the vicinity, and an account of the excavations of 2002. It includes twenty colour plates, several showing the part of the mosaic uncovered last year. Perhaps the most interesting image is of the geophysical survey, which shows the ground plans of the pair of near-identical villas with unusual clarity. The booklet is published by Ex Libris. If you would like a copy, send a cheque for £5 (this includes p & p) to Ex Libris Press, 1 The Shambles, Bradford on Avon, Wiltshire BA15 1JS, or it can be ordered on-line and paid for by credit or debit card at

Danegate, Lincoln

Excavations in Danegate, Lincoln revealed a chequered pavement—this will appear in an appendix to Volume IV of the corpus of Romano-British mosaics. There has been a certain amount of controversy (Lincolnshire Echo, 3 July 2003) because only those parts of the Roman house that would be disturbed by building work (for instance lift shafts) were excavated. The mosaic came from the end of a porticus where it turned a right angle to front another wing. The colours of the chequers were red and cream and red bands flanked the panel. Incidentally, the excavations were in advance of the construction of the long-awaited new City and County Museum (costing £10.5 million).

Also near Market Rasen, Lincolnshire, metal detectorists have found, amongst other things, ‘mosaic tiles’ at a Roman temple site, but we have no further information.

Wymondham, Leicestershire

In excavations directed by Roger Wilson, two new mosaics have been found, both executed in coarse tesserae and both in a fragmentary condition as a result of plough damage. In the porticus is a pavement of cream and blue-grey chequers. In one of the rooms is a cream grid on a red ground, with small red squares at the intersections; it has a broad cream border. On the western side it appears to have been sealed by a tessellated pavement.

Silk Willoughby, Lincolnshire

One of the main problems with producing a corpus of Romano-British mosaics is that new discoveries make it incomplete and out of date as soon as it is published; this is why there will be appendices to Volume IV to include subsequent discoveries too late for inclusion in the first three volumes. Already there are at least five new mosaics in the area covered by Volume I! It is ironic that one of the ‘new’ mosaics was discovered directly as a result of the Corpus. It concerns Mosaic I, 61.1 from Silk Willoughby, Lincolnshire. The entry reads: ‘It is recorded that a mosaic was found a few years ago at Grange Farm when a hole was dug for a cattle grid but no further details are known.’ This passage inspired the search for the mosaic, and indeed part of it was discovered close to the cattle grid. It has blue-grey bands around a ‘make-up’ panel with fragments of a possible running-peltae pattern.

Hawkesbury, Gloucestershire

In July, excavations close to the nature reserve at Lower Woods Lodge by Richard Osgood, Archaeological Officer of South Gloucestershire Council, have revealed part of a mosaic. The geophysical survey shows the site to be a courtyard villa, the main residence being a probable winged-corridor building with pavilions possibly at all four corners. Thus it seems to have two façades; one south-facing; the other facing the courtyard, which is enclosed by a wall with a gateway on the east side and a large, perhaps agricultural, building to the north. The mosaic paves the south-east pavilion. The northern end was excavated and only fragments survive. There seems to be a narrow rectangular panel at the northern end, comprising a red band framing a row of superposed triangles (similar to the border of a second-century mosaic from Wadfield, Gloucestershire). Of the presumed central panel there are traces of a guilloche frame and swastika-meander but most of it is lost. The mosaic has a border of coarse brownish-purple pennant stone, in which there are two small curious features in white, at least one curvilinear. These are not random use of an alternate colour as occasionally occurs but a deliberate motif. Unfortunately very little of either survives; it is conceivably a signature or trademark. The mosaic suffered heavily in antiquity as metal-smelting went on in the room at a later period, causing burnt areas of mosaic, and postholes cut the pavement as well. It is hoped that more of this mosaic will be revealed next year. The pottery appears to be largely second and third century. Another important site has been found to the south at Horton but it has not yielded a mosaic.

Badminton Park, Gloucestershire

Last year, the interest engendered by the discovery of the magnificent mosaic at Lopen, Somerset, led directly to the investigation in a field at nearby Dinnington, where a farmer had come across tesserae and a fine fragment. A similar scenario was played out in South Gloucestershire as a result of the find from Hawkesbury described above. The Duke of Beaufort visited the site and showed great interest, resulting in his giving permission for a geophysical survey to locate the villa at Badminton Park, where a mosaic was reported in the seventeenth century. The survey revealed the plan of a villa and courtyard, and a trial trench was dug in August, exposing two mosaics. One in what is almost certainly the porticus is relatively simple, comprising a coarse purplish-brown Pennant stone border with a band of white about 0.60 m wide running down its centre and relieved by blue double fillets at the margins. The neighbouring room with an apse has a mosaic, which, although ambitious in concept, is relatively crude in execution, although the geometry, while unusual, is accurate. It has a large pair of interlaced squares in guilloche at the centre, tangent to parts of others at the margins, but, unusually, in the corners of the square frame around it there are oblique and crossing bands of guilloche to form kite-shapes, ineptly filled with debased swastikas, etc. The apse has a half-flower central to the cord at the centre from which radiate lines up to the coarse Pennant stone border (about 1.20 m wide); this possibly indicates the position of a semicircular couch (stibadium).

Swalcliffe, Oxfordshire

Edward Shawyer has resumed his excavations at this site in late August. In Room 9 the mosaic was found to have survived in a better condition than those already discovered. (For a summary of previous work here see Mosaic 27 (2000) 10–11). The new mosaic is very similar to that at Llantwit Major both in scheme and motifs, being an arrangement of octagons, squares, rectangles and lozenges. In the central octagon a cantharus partially survives. Its neck is unusual decorated by a band of meander, and is identical to the one found last year only about 12 km north-west at Pillerton Priors, Warwickshire.

Keynsham Roman Villa

When the ARA visited Keynsham cemetery, the site of one of the finest Roman villas in Britain, the door to the mortuary chapel was open and the delegates who peeped inside noticed that the floor boards had been removed, revealing patches of tessellation. Many thanks go to those who let me know. Charles Browne tells me that the chapel is suffering from damp—hence the start of remedial work—and informed the proper authorities. I have been assured by Bob Sydes (the archaeological officer responsible) that the floor will be properly examined before work commences and, if the proposed installation of French drains will disturb any area, archaeologists will excavate it first.

Mosaics and the plough

ARA The Bulletin of the Association for Roman Archaeology, Issue 14 for March 2003, included a well-illustrated article by David Sabin on the villa and mosaic found last year at Pillerton Priors, Warwickshire. Figure 1 dramatically shows scoring across the pavement caused by ploughing, and the continuation of the lines across the field beyond demonstrates that this damage was very recent. The plough has caused untold destruction to ancient sites, and mosaics in particular are extremely vulnerable. In fact, many mosaics were discovered when the plough broke through them and turned up tesserae. A glance through Volume I of the corpus of Romano-British mosaics shows several examples with obvious plough damage (for example, Stanwick and Rudston). Although this destruction has been going on for centuries, modern technology has resulted in deeper ploughing and an even greater risk of loss. Therefore, we can only applaud English Heritage for mounting a campaign to protect archaeology from ploughing with their leaflet Ripping up History. English Heritage is calling for changes in the law, which at present aims to protect the site by scheduling it but cannot always prevent its destruction through cultivation. The leaflet can be printed off from the English Heritage website (it's not easy to find: click on 'Public Policy', then 'Countryside Policy', then 'Farming Policy', or try searching for 'ripping'; the actual address of the leaflet is Alternatively, a copy of the leaflet (no. 50791) can be obtained from English Heritage Customer Services, PO Box 569, Swindon SN2 2YP, tel. 0870 3331181.

Current World Archaeology

Those of you who take the magazine Current Archaeology, will know that Andrew Selkirk is going global. Current Archaeology, issued six times a year, is an excellent way to keep up with British archaeological finds and theories, and now with a relatively little extra subscription, one can take Current World Archaeology in the intervening month. The new magazine is launched in September: see

If you would like to receive the full version of these newsletters regularly, why not join ASPROM?

Newsletter 43 was written by Stephen R. Cosh.
This page is maintained by Ruth Westgate.