Vetusta Monumenta: Ancient Monuments, a Digital Edition

Plate 2.44: Roman Pavements Found at Cirencester and Woodchester


Engraved by James Basire (1730-1802), Plate 2.44 depicts two fragments of Roman mosaics located in Cirencester and Woodchester, Gloustershire. Though earlier drawings are known to exist for at least the Woodchester pavement, these were determined to be inaccurate by the Society and were not used as the basis for their print (Lysons 1817c, 2). Rather, Basire modeled his engraving after images by the antiquarian Samuel Lysons (1763-1819), who drew and colored them in 1787. Basire’s finished plate was published 23 April 1788.


The larger fragment on the plate’s left-hand side represents a section of a pavement found at Cirencester—one of six that have been unearthed near Dyer Street. It features a scattering of marine creatures, a portion of the surrounding outer border, and possibly the wheel of a deity’s chariot in the upper right-hand corner. Unearthed in 1783, it is thought to date from the second century, but its original dimensions are currently unknown (Cosh and Neal 2010, 106).

The second, smaller fragment on the right shows a portion of what was later deemed the “Great Pavement” at Woodchester. First discovered in 1693, the original mosaic panel measured 14.86 meters (48 feet 10 inches) squared, making it the largest known mosaic from Roman Britain. The section depicted on Plate 2.44 was unearthed around 1786 and displays a cross-section of three concentric rings, each demarcated by a triple band of guilloche: the inner contains four birds, the central houses an elephant, while the outer harbors a swirling pattern of acanthus scroll. Orpheus playing his lyre is the mosaic’s primary figure, and though not part of the fragment depicted in Vetusta Monumenta, he would have been placed off-center, allowing for a fountain to sit at the mosaic’s axial point.


No. I. Part of a Roman tessellated Pavement found at Cirencester in Gloucestershire, 1783.

Scale of Feet.

No. II. Part of a Roman tessellated Pavement found at Woodchester in Glouc-rshire.

Scale of Feet.

Paint. Lysons del. / Sumptibus. Soc. Antiquar. Londini. Publish’d according to Act of Parliament, 23d. April 1788. / J. Basire Sc.

Original Explanatory Account: Click here to read the original explanatory account for Plate 2.44.

Commentary by Yoonjae Shin:

The Pavement at Cirencester

In 1783, while excavations were underway at No. 50 Dyer Street, workers unearthed part of an ornate Roman pavement. This would be the first of six mosaics to be uncovered along this roadway, all of which are now believed to have been housed in the same large building (Cosh and Neal 2010, 105). The fragment was uncovered, as the Society of Antiquaries of London (SAL) explain in the first page of their own 1788 explanatory account, “in digging the foundation of a house belonging to John Smith” and found lying “at the depth of six feet below the level of the street.” Once exposed, the mosaic was left open to both viewers and the elements for at least sixty years, allowing antiquarians such as Samuel Lysons to draw what had been uncovered (Cosh and Neal 2010, 106). While this seemingly careless treatment was not particularly uncommon at the time, it likely explains why so little of the pavement survives into the present. The sole piece of the mosaic that exists today is the result of a small excavation which took place in the early 1970s when a lift was being installed in the building that stands where the pavement is buried. This resulted in an eight inch strip of the pavement—“enough to confirm the accuracy of Lysons’ illustration”—being exposed and lifted (Rainey 1973, 48). It is currently in private hands.

Little can be said with much certainty about this mosaic. While more precise dating is difficult, it is generally thought to have been laid sometime during the second century AD due to “its quality and its proximity to mosaics more certainly of that date” (Cosh and Neal 2010, 106). Likewise, the exact dimensions of the pavement are unknown. At least one nineteenth-century writer was willing to speculate, however, that it was “probably eighteen feet square, of which rather more than a fourth part was preserved” (“Tessellated Pavement” 1849, 358). Despite the uncertainties that surround the pavement, contemporary scholars agree that it was likely “a large square marine panel, perhaps originally with Neptune at its centre” and would thus be depicting the Triumph of Neptune (Cosh and Neal 2010, 106). The animals shown in Vetusta Monumenta include a sea-leopard and a sea-horse, surrounded by smaller figures such as fish, clams, scallops, and what Thomas Morgan identified, in 1886, as a “conger-eel,” in blues and reds (80). Two human figures are also visible in this engraving: in the uppermost corner is “a winged cupid apparently swimming beside a wheel (possibly of Neptune’s chariot)” placed above “a second winged cupid, rather damaged, astride a dolphin, clinging on with his right hand and raising his left” (Cosh and Neal 2010, 106). All of this takes place in front of a white background.

Later antiquarians were certainly aware of this illustration as well as others that existed of this pavement. An article in the Gentleman’s Magazine noted that this particular pavement “is engraved in the Vetusta Monumenta of the Society of Antiquaries of London, [...] and more carefully in Lyson’s Reliquiae Britannico-Romanae” (“Tessellated Pavement” 1849, 358). Indeed, curiously, the Vetusta Monumenta engraving lacks features found in other illustrations, notably Lysons’s own Reliquiae Britannico-Romanae from a quarter-century later (1817b, Plate VII). Extending beyond the borders of the Vetusta Monumenta depiction, Lysons’s later image includes more marine creatures along the bottom as well as the second winged cupid in a less damaged state and a small hand, likely belonging to a third cupid, on the right. On the left, part of a Nereid, entangled with an aquatic creature of which only the tail remains, is visible. From Lysons’s later engraving, it is possible to speculate that this panel was framed by

a double fillet and a double latchkey-meander, although only the damaged portion on one side survives. Traces of the contents of two small square compartments indicate that they are filled by a simple eight-petalled flower and an adapted chessboard pattern comprising a diagonal arrangement of staggered squares, forming a quincunx, with triangles in the corners. This border is executed in larger tesserae than the main panel. (Cosh and Neal 2010, 106)

Aquatic scenes such as this were not rare in Roman Britain. Anne Rainey explains that “besides producing a number of very striking geometric designs, mosaicists in Britain drew quite widely on Graeco-Roman mythology for their subjects” and that “marine compositions with Neptune, fishes and dolphins were a particularly popular theme” (Rainey 1973, 15). However, others have noted that the depiction of gods and cupids in such scenes is less common in Britain (Cosh and Neal 2010, 106). Yet it is not just the inclusion of deities that makes this mosaic a fascinating find. Rather, the similarity of this pavement to others uncovered throughout Europe makes this an object of interest. The array of marine creatures backed by white “bears a greater resemblance to similar mosaics found in the Mediterranean area than anything so far discovered in Britain” (Rainey 1973, 47). Even Lysons could not help but make this connection: he follows his depiction of the fragment in his Reliquiae Britannico-Romanae with an outline of a marine-themed mosaic situated in Rome, noting the similarities in content and in style between the two pavements (1817b, Plate VIII no. 2). Through this pavement and others, the connection between Romano-British mosaics and those found in other parts of Europe was becoming more evident.

The “Great Pavement” at Woodchester

While the year 1783 marked the discovery of mosaics near Dyer Street in Cirencester, 1786 saw the recovery of a section of what Lysons would later deem the “Great Pavement” in nearby Woodchester. Perhaps making this proximity more apparent, one writer noted that the town of Woodchester “may be found, in the map of Gloucestershire, at the intersection of two lines; one drawn directly north and south through Gloucester, and the other due east and west, rather less than a mile north of Cirencester” ("An Account of Roman Antiquities" 1798, 2-3). Perhaps due to the nearness of these two locations, the mosaic fragment found here was published alongside the one unearthed at Cirencester. (This follows a precedent set by Plate 2.9, where the mosaics found independantly at Winterton and Roxby were engraved on one plate.)

Likely originating from a palatial villa, the Great Pavement was first discovered below a churchyard in 1693 (Rainey 1973, 163; Cosh and Neal 2010, 214). The earliest mention of this pavement, Lysons notes, “is in the Additions to Camden’s Britannia, published in 1695, by bishop Gibson” (1817c, 2). However, the first representation of it would not come for another two decades. The earliest illustrations of any part of the mosaic are those by Edmund Brown, who exposed and drew part of the mosaic in 1712. Richard Bradley’s drawings—one of which mirrors the image found in Vetusta Monumenta—came a decade later (Cosh and Neal 2010, 214). A portion of the pavement, that first drawn by Brown, was again uncovered around 1786 “by the Rev. Peter Hawker, rector of Woodchester, who has taken all possible care for its preservation,” as noted in the Society’s explanatory account. By the time Lysons himself wrote about these findings at the beginning of the 19th century, however, this section was no longer extant: “though much care was taken by the rector for its preservation, the wet and frost have long since destroyed it” (Lysons 1817c, 2).

Though earlier illustrations of this pavement existed, they were not used as the basis for the Society’s print. Not all of Bradley’s illustrations had survived, and those that had are said by Lysons to be “very inadequately drawn” (Lysons 1817c, 2). Conversely, the Society’s account notes that Brown’s drawing was “very small, and without a scale” (Gough 1788, 1).1 Because of the dissatisfaction of both Samuel Lysons and the SAL, Basire instead modeled his engraving after new images by Lysons himself, who drew and colored them in 1787. Lysons’s drawing for Vetusta Monumenta shows a cross-section of three distinct sections. The innermost section contains various birds and trees and is surrounded by an outer border of laurel wreath. The center section, flanked by three-stranded guilloche, depicts an elephant standing in front of a small tree. The elephant is decorated with lattice, and while Cosh and Neal conjecture that this could be “artistic licence by the copyist,” they note that “the tree behind the elephant, the birds and other patterns are very similar to surviving examples, and therefore we must assume that the engraving is a fairly faithful rendering” (2010, 220). Another point of conjecture is the elephant’s size: though the SAL’s engraving gives the elephant a length of nearly three feet long, Lysons notes that most of the central animal figures “are about four feet in length” (Lysons 1817c, 3). The outer section contains rolling acanthus scrolls, and another outer border of triple-stranded guilloche is just discernible. Beyond that, between the mosaic’s outermost square border and the concentric main area, are figures which Lysons identified as Naiads carrying urns (Lysons 1817c, 2).

Lysons’s work on the Woodchester pavement would not end with the depiction of this fragment. He reports that in 1793, while a vault was being dug for the burial of a John Wade of Pudhill, “so considerable a portion of the same pavement was laid open, as, together with other openings,” which were made in the course of that and the following year, allowed one to establish its form and dimensions—namely, “a square of forty-eight feet ten inches” (Lysons 1817c, 2).2 From his later observations, Lysons could conclude that “the figures of a gryphon, a bear, a leopard, a stag, a tigress, a lion, and a lioness are now remaining,” but by the time of his writing, the figures “of a boar and a dog, which are to be seen in Mr. Brown’s drawing, together with that of an elephant, have since been destroyed, and no part now remains of the two others necessary to fill up the whole space.” He was able to conclude “from the memorandum, written on the margin of one of Bradley’s drawings,” that the innermost section of the mosaic, a central octagon, “contained figures of fish, and that ‘about the centre there was a starlike figure’” (Lysons 1817c, 3).

The central figure of the mosaic “represents Orpheus playing on the lyre, which he rests on his left knee,” a figure that sits curiously within the avian section of the mosaic rather than at its center (Lysons 1817c, 3). Beneath the mosaic is a hypocaust, a kind of underfloor heating system (cf. Plate 1.57) that is unusual for rooms of a circular shape (Cosh and Neal 2010, 214). However, channels from the north wall to the mosaic’s center are absent from the hypocaust’s layout. Cosh and Neal speculate that, based on scarring in this same location on the pavement, this may have been where lead piping was laid instead, to feed a fountain which would have stood at the mosaic’s center (2010, 214-17). This would explain Orpheus’s skewed placement within the pavement’s structure. A mosaic of this size must have consisted of a considerable number of tesserae, and Lysons speculates that “the whole when entire could not therefore have contained less than a million and a half of them.” He also surmises that these tesserae, “for the most part, nearly cubes of half an inch,” were made from resources natural to Britain except, notably, for the white tesserae, whose hardness suggested to him a resemblance to Palombino marble from Italy (Lysons 1817c, 4).

Dating, based on style alone, places this mosaic sometime in the fourth century (Cosh and Neal 2010, 212). Based on dating and on similarities in themes, patterns, and idiosyncratic embellishments within geographical areas, David Smith initially organized four different groupings or “schools” of mosaics in the 1950s—one of which was the “Corinian” school, based at Cirencester. He would eventually break these down into “Corinian Orpheus” and “Corinian Saltire” schools (Cosh and Neal 2002, 21). Smith’s groupings indicated that groups of craftsmen often worked within distinct regional areas, even when their ideas travelled or were emulated elsewhere. Nonetheless, despite the existence of distinct groups of craftsmen, Cosh and Neal note that “it is readily noticeable that the ambulatory around the Orpheus panel at Woodchester differs in style, which hints at collaboration between groups of mosaicists on large contracts” (2002, 26). Indeed Woodchester would fit this description: the site consisted of a major villa containing around 64 rooms, and the Great Pavement “is the largest mosaic in Britain, measuring just under 15m square” (Cosh and Neal 2010, 214). Yet it is not just the mosaic itself that is great: “the number of mosaics at Woodchester also appears to be exceptional, for the 13 or 14 in situ are only exceeded in Britain by Fishbourne” (Clarke et al. 1982, 206). The knowledge of a few of these mosaics—along with the discovery of additional hypocausts, statues, coins, and pottery at the same site—led Lysons to conclude that the Woodchester pavement was constructed not for a private individual, but for a Roman personage of high importance. In that case, he surmises that the building in which the pavement was found may have served as the residence “of the Proprætor, or, at least, of the governor of this part of the province, and occasionally, perhaps, of the Emperor himself” (Lysons 1817c, 17). While Cosh and Neal do not go to quite these same lengths, they do agree with his general assessment and suppose that Woodchester Villa “must have been the residence of a wealthy and important individual” (2010, 105).

Sections of the mosaic “continued to be cut by graves” until the end of the nineteenth century. About forty-two percent of the pavement was surviving when it was recorded in situ during its last exposure in 1973. Additionally during this time, part of the outer circle was lifted; this section is now held by the British Museum. The rest of the pavement was reburied and currently lies beneath the churchyard at Woodchester (Cosh and Neal 2010, 214). It may not be uncovered for the foreseeable future: the last time it was uncovered in 1973—for a total of seven weeks—the over 140,000 visitors who came to view the mosaic caused such traffic congestion that the local populace decided that “it should never be unearthed again” (Clarke et al. 1982, 197; Gloucestershire Museums 2010).

Reception: Lysons, Britain, and the Roman Empire

While the interest in mosaics was already on the rise—reflected in the increase of mosaics being printed in Vetusta Monumenta and elsewhere—Martin Henig claims that “the culmination of the interest in and investigation of mosaics came at the end of the eighteenth century and the beginning of the nineteenth, with the gentleman-scholar Samuel Lysons” (1995, 178). In fact, he notes that “Lysons’ excavations and the Reliquiae Britannico-Romanae, however imperfectly realized, do seem to have broken new ground in the study of the mosaic art of Roman Britain” (179). Despite any advancements he may have made in the field of archaeology (Sweet 2013), Lysons was also part of the vanguard changing the reception of Romano-British pavements. Particularly noteworthy are his efforts in recontextualizing Britain’s Roman past.

Leading into the nineteenth century, Britain’s role in the Roman Empire was not well regarded or understood. Sarah Scott notes that “early antiquarian interest in Roman Britain,” largely perceived as a military outpost and cultural backwater, “focused on the excavation and recording of military remains and roads” (Scott 2013, 5). Indeed, this perception would even color the approach to non-military Roman antiquities. For example, a 1760 article discussing the mosaic found at Winterton, Lincolnshire (Plate 2.9) suggests that the pavement had a strictly utilitarian, military purpose: flooring the tents of Rome’s military leaders. Based on such coarse evaluations, then, it would appear that Britain was not a particularly dazzling jewel in Rome’s crown. Thus Lysons and his contemporaries, those who studied local antiquities and highlighted their historical significance, stimulated pride and interest in the artifacts and helped to usher in an “emergence of a real sense of ‘Englishness’” (Scott 2014, 301). One could look back at Britain’s Roman past with both aesthetic pleasure and cultural pride.

Yet, while he was certainly interested in British history, Lyson’s broader aim was not just to solidify further a sense of national identity. Rather, he strove to give Britain a place in the larger scale of the Roman Empire, and the finds at both Cirencester and Woodchester would help Lysons to reshape the popular conception of Roman Britain. Since the 1700s, a number of Roman artifacts had been found in Cirencester, and recent archaeological data suggest that the Roman town found here was a civitas—the center of a self-governed political unit which was oftentimes based on a pre-existing tribal territory (Darvill and Gerrard 1994, 57). With the SAL highlighting in their explanatory account the number of recent finds in the area—“Dyer-street, which is one of the four principal streets of Cirencester, leads to the Foss-way, one of the Roman roads, which met there; and tesselated pavements have frequently been found on both sides of the street” (Gough 1788, 1)—other nineteenth century commentators seem to have implicitly understood the importance of the site. As Scott observes, “throughout the course of the eighteenth century the discovery of increasing numbers of artefacts of non-military origin began to show that Roman Britain had been an established and civilized part of the Roman Empire” (2014, 303).  The sites at Cirencester and Woodchester aided antiquarians in backing this claim.

Likewise, Lysons would write of the Woodchester pavement that “the complete design of this pavement [...] is, I believe, equaled by few of those discovered in other provinces of the Roman Empire, and is undoubtedly superior to any thing of the same kind hitherto found in this country” (Lysons 1817c, 2). The existence of these beautiful mosaics hinted at the existence of equally impressive villas to house them. In the advertisement to his Reliquiae Britannico-Romanae, Lysons wrote that “although scarcely any traces exist above ground, of the building erected by the Romans, during the four hundred years they continued in possession of this Island [...] perhaps it is not too much to say, that no province of the Roman Empire contained a greater number of extensive and richly-decorated villas” (Lysons 1817a, iii). The scale and magnificence of these two pavements—one found in the heart of a civitas and the other among the largest mosaics from Roman Britain—helped generate an interpretation of Britain’s Roman past that highlighted the country’s place of importance in relation to Rome.

Lysons’s reevaluation of Roman Britain was not merely an aesthetic exercise, however. Scott writes that Lysons was “at the heart of an intellectual elite which genuinely believed that the encouragement and promotion of British intellectual and artistic achievement was an essential undertaking in the national interest” (2014, 304). During the period in which the SAL published its engraving of the Cirencester and Woodchester pavements—bookended by the Anglo-French War (1778-1783) and the French Revolutionary Wars (1792-1815)—a growing sense of nationalism was beginning to develop across Europe. In England, this paradoxically created “a strong desire to be seen to live up to common European standards in collecting and connoisseurship against the backdrop of a shared classical heritage[...]; in particular, there was a concern to challenge the widely held view that Britain was culturally inferior to its Continental neighbours, most notably France” (Scott 2014, 294-95). Lysons’s methods, depictions, and descriptions of British mosaics set a new benchmark for antiquarian work in England. They also made a point to readers and scholars abroad, one whose nationalist sentiments would take on political undertones in the 19th and 20th centuries.

Some of Lyson’s major texts, such as the Reliquiae Britannico-Romanae, were aimed at both a domestic and a foreign audience and “asserted British cultural achievements under Roman rule while showcasing British skills in the recording and interpretation of antiquities” (Scott 2013, 16). Thus, while he wished that the precise information within the tomes would warrant them to be “an acceptable addition to the Libraries of [England] and other countries,” he also challenged his readers to include Britain in their comparative studies of history. Within his work, Lysons often drew upon his extensive knowledge, classical and modern, to argue the similarities “between British and Continental discoveries through reference to examples discussed in contemporary European publications and in well-known sources such as Montfaucon, de Caylus, and Winckelmann” (Scott 2014, 312-13). By showing how ideas found in Europe were mirrored in British pavements, Lysons firmly placed Britain in a greater, broadly European, classical context. Therefore, while Scott argues that Lysons’s comparisons are not always the most apt, she maintains that “it is nevertheless significant that Lysons was making comparisons and was aware of the potential of this kind of analysis” (2013, 12). With his laudatory approach to Romano-British artworks as much as his exacting methodology, Lysons helped champion a reconsideration of Britain’s classical past.


Martin Henig claims that “the tradition of high-quality mosaic-recording and publication did not long survive and its decline is signaled by Sir Richard Colt Hoare in his treatment of the ‘Roman Aera’, appended in 1821 to Auncient Wiltescire” (1995, 180). If this claim is true, then the SAL’s engraving captures an intriguing moment in British historiography. Interest in Roman pavements was at a high point—some 30 years before its supposed decline—with Lysons, the SAL, and other antiquarian bodies producing numerous illustrated publications on mosaics during this time. This was all part of an enthusiasm for rethinking Britain’s historical claim to art and culture as part of Rome’s empire, spurred on by continuing discovery in England and changing circumstances in Europe. The rehabilitation of Roman Britain, begun by antiquarians and their search for “Englishness,” would thus culminate in the knowledge of the island’s importance to Rome.

Thus, the Roman mosaics engraved by the Society can be seen to fit into a larger national project, deepening the understanding of Britain’s classical past. The debate about British identity and Roman national origins would not disappear alongside the enthusiasm for engraving and printing such antiquities. Instead, it would come to the fore even more prominently during the nineteenth and twentieth centuries, in works such as Francis Haverfield’s The Romanization of Roman Britain—where he claimed that the emergence of orderly civilization “was the work of the [Roman] Empire,” a task that Britain would later claim to take up itself (Haverfield 1905, 2). From this vantage point, the SAL’s plate on Cirencester and Woodchester might be viewed as crystallizing the turning point in Britain’s understanding of itself, a harbinger of academic inquiry to come. The society’s engraving, then, captures not only two fragments of mosaics now lost but also an historical moment contributing to Britain’s changing self-definition.


[1] A copy of Brown’s drawing (by Priscilla Combe) is viewable at the British Library:

[2] A print based on Lysons’s full reconstruction is viewable at the British Museum:

Works Cited:

“An Account of Roman Antiquities Discovered at Woodchester, in the County of Gloucester. By Samuel Lysons.” 1798. The British Critic: 1-7.

Buckman, James and Charles Newmarch. 1850. Illustrations of the Remains of Roman Art, in Cirencester, the Site of Antient Corinium. London: G. Bell.

Clarke, Giles, Valerie Rigby, and John D. Shepard. 1982. “The Roman Villa at Woodchester.” Britannia 13: 197-228.

Cosh, Stephen R. and David S. Neal. 2002. Roman Mosaics of Britain, Volume 1: Northern Britain. London: Illuminata Publishers for the Society of Antiquaries.

------. 2010. Roman Mosaics of Britain, Volume 4: Western Britain. London: Illuminata Publishers for the Society of Antiquaries.

Darvill, Timothy and Christopher Gerrard. 1994. Cirencester: Town and Landscape: An Urban Archaeological Assessment. Gloucestershire: Alan Sutton Publishing Ltd.

Gloucestershire Museums. 2010. “Orpheus Pavement from Woodchester.” A History of the World in 100 Objects. BBC,

Gough, Richard. 1788. “Vol. II. Plate XLIV.” Vetusta Monumenta, vol. 2.

Haverfield, Francis. (1906) 1912. The Romanization of Roman Britain. 2nd edn. London: Oxford University Press.

Henig, Martin. 1995. The Art of Roman Britain. Ann Arbor: The University of Michigan Press.

Lysons, Samuel. 1817a. Reliquiae Britannico-Romanae, vol. 1. London: Printed by T. Bensley.

------. 1817b. Reliquiae Britannico-Romanae, vol. 2. London: Printed by T. Bensley.

------. 1817c. Reliquiae Britannico-Romanae, vol. 4. London: Printed by T. Bensley.

Morgan, Thomas. 1886. Romano-British Pavements: A History of Their Discovery and a Record and Interpretation of Their Designs. London: Whiting and Co.

Rainey, Anne. 1973. Mosaics in Roman Britain. Totowa: Rowman and Littlefield.

Scott, Sarah. 2013. “Samuel Lysons and His Circle: Art, Science and the Remains of Roman Britain.” Bulletin of the History of Archaeology 23, no. 2: 1-22.

------. 2014. “Britain in the Classical World: Samuel Lysons and the Art of Roman Britain.” Classical Receptions Journal 6, no. 2: 294-337.

“Tessellated Pavement Found at Cirencester.” 1849. Gentleman’s Magazine 32: 357-60.

This page references: