This page has paths:
- 1 2018-05-15T18:32:48-05:00 Craig Dietrich 2d66800a3e5a1eaee3a9ca2f91f391c8a6893490 Volume 1, Plates 27 — 49 Crystal B. Lake 3 plain 2018-08-29T17:52:21-05:00 Crystal B. Lake b7829cc6981c2837dafd356811d9393ab4d81adc
This page is referenced by:
Plate 1.48: Roman Pavement Found at Cotterstock
Scholarly Commentary with DZI View for Vetusta Monumenta, Plate 1.48. Commentary by Elizabeth J. Hornbeck.
Plate: Plate 1.48 shows a Roman mosaic floor discovered in July 1736 in the parish of Cotterstock in Northamptonshire, between the villages of Cotterstock and Glapthorn. Its discovery attracted the attention of George Lynn (1676-1742), who then brought it to the attention of William Stukeley (1687-1765) and the Society of Antiquaries of London; a detailed drawing was made later that year by William Bogdani (1699-1771), son of the painter Jacob Bogdani, based on measured drawings by Lynn and his son George. It was engraved by George Vertue (1684-1756) for the Society in 1737. The print occupies a two-page spread, the second of only six plates in Volume I to do so. The engraving omits the bottom quarter of the mosaic floor to make room for the inscription, allowing the most intricate portion of the mosaic (the central 10 x 10 foot block) to be shown at a larger scale than would have been permitted if all four sides of the border were included. The engraving as well as the drawings on which it was based show a reconstructed image of a pavement that was partially lost over centuries of plowing the field where it was discovered.
Object: When it was found in 1736, the surviving section of the mosaic pavement measured 19 feet 11 inches by 14 feet 11 inches. Due to the large number of loose tesserae found in the plowed field, as well as to a demand for symmetry, Stukeley and Lynn surmised that one side of the original pavement had been destroyed over time by plowing, and that originally it was approximately 20 feet square. It was a multicolored mosaic consisting of blue, red, light yellow, and white tesserae (SAL Minutes III.16-17). The mosaic floor discovered in 1736 was part of an enormous Roman villa which, however, has not been excavated, and is known primarily through modern surveying techniques. The mosaic floor, having been uncovered twice in 1736, was reburied and is no longer visible. Based on comparisons with securely dated mosaics from a Roman villa at Stanwick, Neal and Cosh hypothesize that the Cotterstock mosaic dates from about AD 375 (Neal and Cosh 2002, 233).
EXIMIUM HOC OPUS TESSELLATUM / Romanae apud BRITANNOS Magnificentiae Monumentum, juxta COTTERSTOCK in Agro Northamtoniensi, mense Augusto MDCCXXXVI repertum, ne Vulgi Profani temeritate Posteris perditum esset, Georgius Lynn Pater & Filius de Southwick villâ proxime vicinâ, et Gulielm.us Bogdani Armigeri delineaverunt; / SOCIETASQ. ANTIQUARIA Lond: Suis Sumptib.us in Aere incidi curavit An.o Dom. MDCCXXXVII
This [is an] extraordinary tessellated mosaic work. [It is] a monument of Roman magnificence in Britain, and it was uncovered near Cotterstock in the county of Northampton, in the month of August 1736. George Lynn, father and son, from the neighboring house in Southwick, and William Bogdani, Esquire, drew it, so that it would not be lost to posterity on account of the unlearned public’s carelessness. The Society of Antiquaries, London undertook its engraving in copper at its own expense, in the Year of our Lord, 1737.
Commentary by Elizabeth J. Hornbeck:
In July 1736 a large, elegant, and colorful mosaic floor was discovered in the parish of Cotterstock in Northamptonshire, between the villages of Cotterstock and Glapthorn, by a farmer who was plowing a village field. According to a 1737 report in the Northampton Mercury the plow had turned up “several little stones which made a very uncommon appearance,” and which upon further investigation were believed to be tesserae from a long-buried Roman mosaic floor (quoted in Upex 2001, 89). This mosaic almost immediately attracted the attention of antiquaries, including George Lynn, who lived at Southwick Hall, less than a mile north of the site, and William Stukeley, who lived in Stamford in Lincolnshire, about twelve miles from Cotterstock.
Stukeley describes the site that he encountered on 28 August 1736 when he visited it at Lynn’s invitation. Lynn, Stukeley, and a Mr. Laurence “got workmen, and cleared the Roman pavement above mentioned, and took an exact drawing of it, with all its dimensions and colours” (Stukeley 1887, 49). He reports that, due to repeated plowing of the land, one end of the pavement was entirely plowed up, its tesserae being sprinkled over the whole field. Stukeley and his colleagues “found two-thirds of the work intire, and so much of the remainder as enabled us by symmetry to draw out the whole, except the central part of the upper compartment. Therefore to prevent deformity, we judged proper to fill it up by the same central work as the other, turning it the contrary way, particularly the heart-like work” (51). Thus the drawings and the engraving based on them are restored images of an imperfectly surviving mosaic pavement. Stukeley records the overall dimensions of the surviving pavement as thirty Roman feet by twenty, adding: “The beautiful part of the work within is a small matter less than 10 of our English feet, being 10 compleat Roman feet” (51).
Stukeley speculated correctly that the mosaic pavement belonged to a Roman villa, but the villa itself has never been excavated. It is known primarily through twentieth-century investigations. Stephen G. Upex (2001) provides a thorough account of the historical record, as well as of modern archaeological investigations. This commentary is deeply indebted to his scholarship.
After the initial discovery of the pavement in July 1736, it was reburied. The owner of the site was only willing to have the area opened once more, which was the occasion for the visit by Stukeley, Laurence, and Lynn for the purpose of making measured drawings. The mosaic floor was again reburied and is no longer visible. It presumably remains in situ, somewhat intact, but with some destruction having occurred when it was exposed in the eighteenth century. The villa as a whole is known to have suffered from subsequent stone robbing and unrecorded antiquarian excavation after its brief period of publicity.
While the mosaic was exposed for the second time, and presumably for a longer period, it was heavily damaged by the large numbers of visitors it attracted. One admirer, the fourth Earl of Cardigan, who lived at Deene Park ten miles to the west, removed a large chunk of the mosaic—about one square yard—and had it taken back to Deene, where it “formed a centrepiece for the floor of a summerhouse at the end of a water-garden” (Upex 2001, 62-63). Upex attributes the Earl’s actions to “a strange and early attempt at conservation and ... preservation.” Neal and Cosh have identified a fragment of a mosaic, found in 1996 in the library of Drayton House, Lowick, as “quite likely” a fragment from the Cotterstock mosaic. The fragment consists of a band of three-strand guilloche and part of a swastika-meander with double returns, as at Cotterstock; it is rare, according to Neal and Cosh, for two mosaics to be so similar (Neal and Cosh 2002, 246). Thus this may well be the fragment removed by the Earl of Cardigan in 1736.
After measuring and drawing the mosaic pavement, Lynn and Stukeley brought the discovery to the attention of the Society of Antiquaries of London (SAL). It is first mentioned in the Society’s minutes on 26 May 1737, when William Bogdani presented a drawing he had made of the mosaic, based on the sketches by Lynn and his son. The SAL voted to have Vertue engrave a print of the drawing, with Lynn’s permission (SAL Minutes III.16-17). At the meeting on 30 June 1737 Lynn gave his consent for the pavement to be engraved, and Vertue was ordered to begin the engraving (III.26-27).
At the meeting on 3 November 1737 Vertue showed the print he had engraved of the image “delineated by George Lynn Esqs. Father & Son, and perfected by Mr. Bogdani” (SAL Minutes III.61). The inscription intended to accompany the plate was recorded in the minutes as well and on 24 November Vertue carried a proof of the inscription for review by SAL members (III.69). Vertue brought the completed copperplate engraving to the meeting on 4 January 1738, when it was ordered that three prints of the Cotterstock mosaic be delivered to each member of the SAL. At the same meeting the members voted to present two dozen prints of the Cotterstock mosaic to Messrs. Lynn Father and son, and one dozen to Mr. Bogdani, “for their care and trouble in delineating the same” (III.80-81). A month later, Lynn proposed that prints of the Cotterstock mosaic be delivered and sold to some gentlemen in the neighborhood to raise money for the SAL; the members voted to deliver 20 prints to Mr. Lynn to be sold for two shillings each (III.99).
It is also worth noting that during this same time, Maurice Johnson (1688-1755) brought to the SAL sketches of bronze instruments and other objects found at Cotterstock that were in the possession of John Hardy, Vicar of Melton Mowbray. These sketches have been transcribed into the minute book; there are sketches of seven objects, including a stylus and a fibula (SAL Minutes III.97). Those objects confirm the amateur digging (largely undocumented) that occurred at sites where antiquities were known to have been found. The minutes from this period also reflect a particular interest in the subject of Roman mosaic pavements: prints and drawings of other pavements were exhibited in the late 1730s and dissertations on the subject were read in the spring of 1739 by James West and a Mr. Bambridge (III.206, 218).
Members of the SAL shared a burgeoning interest in Roman Britain, which increased with the 1732 publication of John Horsley’s Britannia Romana. Bernard Nurse notes an additional reason for the enthusiasm for Roman Britain: in the eighteenth century, Roman remains “were seen as a tangible link between the growing British empire and that of the Romans” (2007a, 127). The study of classical antiquities became the chief interest of the SAL in the middle 1730s (Evans 1956, 93). The Society’s keen interest in the fine mosaic discovered at Cotterstock demonstrates their attention to such archaeological discoveries, which would later be officially signaled by the decision in November 1739 to compile as completely as possible a list of all Roman mosaics found in Britain (Evans 1956, 93). The men who brought the Cotterstock mosaic to the attention of the SAL were part of a close-knit fraternity of antiquarians that included Lynn and Stukeley, who had travelled extensively around England studying ancient monuments, and William Bogdani of Hitchin, who was Clerk of the Ordnance Office in the Tower of London (Evans 1956, 90-91).
These men belonged to a network of antiquaries with Johnson at its center. Johnson lived in Spalding, a market town in Lincolnshire, thirty-two miles northeast of Cotterstock and one hundred miles north of London. He was a founding member of the Gentlemen’s Society of Spalding. This group shared not only an interest in antiquarian and historical studies, but also an interest in supporting “mutual benevolence, and their improvement in the Liberal Sciences and Polite Learning” (Evans 1956, 53-54). Johnson brought his friend Stukeley into the Spalding Society, of which George Lynn was also a member. Both Johnson and Stukeley were to be founder-members of the revived SAL. Vertue and Bogdani, among others, were also members of both societies. Bogdani, who was married to Johnson’s sister, served as acting director of the SAL for one year beginning in January 1738/9 and prepared an index to the minutes of the Spalding Society in the 1740s (90-91). Lynn was also related to Johnson by marriage.
Upex credits Lynn (the father) as being instrumental in recognizing the mosaic’s archaeological significance and in bringing it to the attention of the SAL (2001, 62n18). The presence of a mosaic pavement, and especially one so detailed, multicolored, and finely made, indicates the wealth and status associated with this villa. Johnson’s leadership in the Spalding Society helped to cultivate attention to antiquities and an interest in recording them; he and Bogdani made drawings of Croyland Abbey together in 1738 (Nichols 1790, 91n). Lynn and Bogdani both read mathematical papers at the Spalding Society as well (Nichols 1790 29, 57), and it should be noted that there is a mathematical aspect to the problem of drafting the mosaic with such accuracy that differences in the size of the tesserae used in various parts of the design are captured. The SAL minutes record “an illuminated drawing of part of the Roman pavement in Cotterstock field made all of small dies, of six tenths of an inch square, the border or outer Margine is 5 foot wide, and made of larger dies which are one and 1/4 inch square” (SAL Minutes III.130). In general, Roman mosaics used smaller tesserae for intricate, detailed mosaics, and larger ones for less detailed mosaics, like the simple border that frames the central panel of the Cotterstock mosaic.
Cotterstock Roman Villa
The Cotterstock mosaic was the first recognized evidence of a Roman villa at this location. Besides the drawings of small objects that were brought to the SAL in February 1738, there is no record of any further discoveries at Cotterstock until 1798, when a second, considerably smaller mosaic pavement was discovered and excavated. An engraving was published by William Fowler in 1802, and in 1828, Edmund Artis republished illustrations of both of the Cotterstock mosaics. Artis’s detail of the larger mosaic is almost certainly a copy of Vertue’s engraving, since the mosaic has remained underground since 1736 (Upex 2001, Plate IA).
The elaborate central area of this mosaic is approximately ten by ten feet. Its intricate design incorporates a variety of geometric and abstract patterns arranged symmetrically with red, white, and blue-gray tesserae. At the center is a large poised square containing a flower with four heart-shaped petals with their lower points meeting at a red corolla (circle) in the center. The sides of this poised square are tangent to L-shaped panels framing meanders and guilloches, separated by triangles and lozenges (partial eight-lozenge stars), all set within a larger square frame. This square is enclosed within a second, larger frame consisting of a swastika meander with double returns, and four rectangular frames enclosing simple guilloche patterns. Then there is a third frame containing a three-strand guilloche border. This central design uses smaller tesserae to provide precise details for the intricate patterns. Larger tesserae are used for the border area that surrounds the central design. This border consists of six simple unadorned frames in variegated colors. The overall effect is of a white ground with colored stripes setting off a colorful and intricate central panel that would have been the main focus of the room where it was located.
Throughout the eighteenth century, the area surrounding the villa was under open-field cultivation, though Upex points out that “the outline of the villa was clearly respected by the medieval and post-medieval farmers who ‘boxed in’ the whole building platform area. Furlong 1, which encompasses the main building ranges, may have been created to withdraw it from cultivation because of the underlying stone debris which would have damaged ploughs” (Upex 2001, 60). (The surrounding terrain is clay.) In both 1614 and 1635 the area was known as Stenborrow Field, and in 1815 as Stemborough Field, names meaning “stone hill” or “rock hill.” In 1737 Upex’s Furlong 1 was called “the Guild Acre,” which could refer to gold (or coins) having been found there some time in the past, or possibly to other features of the terrain. The parish was enclosed in 1815, after which the villa site was put under permanent pasture. During Enclosure, several of the villa’s walls were cut through by ditches.
John Pointer (1668-1754) and the History of Roman Mosaics in Britain
Roman mosaics were uncommon but not unknown in Britain in the early eighteenth century. One of the earliest published descriptions of Romano-British mosaics dates from 1713: An Account of a Roman Pavement Lately found at Stunsfield in Oxford-Shire, Prov’d to be 1400 Years Old, by John Pointer of Merton College in Oxford. This book was included in the library of James West (1703-1772), himself a prominent member of the SAL at the time, who took the lead in documenting three Roman mosaic pavements discovered at Wellow in 1737 (Plates 1.50-1.52), just one year after the Cotterstock mosaic was found. Pointer’s short treatise is an invaluable early antiquarian work that both aided the SAL in its own project and helps to document the burgeoning interest in Roman Britain during the early eighteenth century.
Pointer’s forty-page treatise describes, analyzes, and interprets a newly discovered mosaic at Stunsfield, which he calls “the most Elaborate Piece of Roman Workmanship, of this sort and One of the Finest of the Tesselated Pavements, that has hitherto been found in all Britain” (1713, 24). The treatise includes a single illustration labeled “Outlines of the Chief Figures on the Pavement,” inside a roundel. The crude engraving depicts a man holding a cup, and standing (dancing?) in front of a beast. The identities of both have become the topic of much speculation; Pointer himself identifies them as Bacchus and his panther (30). With only a single crude illustration, the treatise relies almost entirely on textual descriptions of the fifteen or so mosaic pavements it mentions. This approach contrasts strongly with Vetusta Monumenta, which relies almost entirely on high quality visual representations, with very little accompanying text. The precision shown in the Vetusta Monumenta plates demonstrates a far greater emphasis on empirical methods for recording and preserving antiquities. Coincidentally, the geometric mosaic at Cotterstock lacks the figures that particularly attracted antiquarian attention to earlier finds including the mosaics at Stunsfield and at Woodchester.
In analyzing the Stunsfield mosaic, Pointer provides an exhaustive history of Romano-British mosaics, beginning with descriptions by Roman authors such as Vitruvius, through antiquarian discoveries in Britain. He begins by mentioning several Roman pavements reported by Robert Plot (1640-1696) in his Natural History of Oxfordshire (1677) (Pointer 1713, 20, 34). Pointer also refers to an account in the Philosophical Transactions of the Royal Society of a mosaic pavement found near All Saints Church in Leicester some forty years previously. In Gibson’s English translation of Camden’s Britannia, Pointer finds references to three pavements found at Ker-Went in Monmouthshire, Wales, in 1689, all of which were destroyed by exposure to harsh weather (34-36).
The same edition of Camden recorded a pavement discovered in 1692 on the estate of one Henry Tomkins of Kaer-Leion in Monmouthshire. Pointer explains that “Mr Tomkins took care to preserve what he cou’d of this valuable part of the Piece of Antiquity, by removing a considerable part of the Floor, in the same order it was found, into his Garden,” which suggests that the actions of the fourth Earl of Cardigan, who removed a large piece of the Cotterstock mosaic to be installed in his summerhouse at Deene Park, were not unprecedented (1713, 36). At the same time, Pointer describes damage to earlier finds such as Bybury, where the tesserae were, “by degrees, all taken away by great numbers of People that came out of Curiosity to see it”—thus illustrating possible grounds for the caution that was exercised in reburying the pavement at Cotterstock (37).
Pointer also mentions several other mosaic pavements found in Northamptonshire prior to 1713 (1713, 37-38). While most of these discoveries were made by farmers plowing fields, Romano-British mosaics were also found in cities, including the one at Leicester and another at Bishop’s Gate in central London, which is described in a letter from Dr. Woodward, of the Royal Society, to Sir Christopher Wren, and included in an eighteenth-century edition of John Leland’s Itinerary (36).
Pointer’s catalogue of Romano-British mosaics also includes references to a vast mosaic at Woodchester that would not be uncovered in its entirety until 1793, when Samuel Lysons excavated it and, in subsequent years, much of the surrounding villa. The mosaic would become known as the Orpheus mosaic which, at thirty-nine feet (twelve meters square), is the largest discovered north of the Alps (Nurse 2007b, 130). In Pointer’s time it was known only through frequent discoveries in a church cemetery:
Pointer’s reference to pavements in “Religious Houses” is ambiguous. His conflation (apparently deliberate) of ancient and modern religious sites suggests an unbroken historical continuity at religious sites that persists under different cultures. His assertion that the Stunsfield mosaic would have been “improper” in an ancient temple is spurious, given that he himself believes it to represent a Roman deity, Bacchus. It is correct, nevertheless, that most or all of the Roman mosaics in Britain belonged to wealthy private villas and not to temples.
There have been likewise such sort of Pavements found in places where Religious Houses have been built, as particularly at Woodchester, in Glocester-Shire, where in the Church-yard, in digging the Graves, they us’d to find such Tesseraick Work of Painted Beasts and Flowers, 2 or 3 Foot deep—as we are inform’d by Dr. Gibson in his Additions to Camden; and Tradition tells us, that there had been a Religious House built there. But our Pavement at Stunsfield cannot be supposed to have belong’d to any Religious House, or Temple, the Figure upon it being so very improper. (Pointer 1713, 36-37)
Confusing the issue, Pointer’s final example of Roman mosaic work is a pavement at Westminster Abbey that was apparently created in the thirteenth century by Roman artisans who were brought to London from Rome in 1260 at the behest of Richard de Ware, Abbot of Westminster (1713, 39). In Pointer’s discussion, the apparent stylistic similarities between the medieval mosaic and ancient examples made them all of one artistic tradition, defined not by historical period but by stylistic consistency. This approach was ultimately abandoned by modern scholarship, but still persists in the work of Richard Gough (1735-1809) at the end of the eighteenth century (Gough 1790).
Overall, Pointer’s text documents the craze for Roman antiquities in Britain that was already in full force in the seventeenth century, as evidenced in the crowds of curiosity-seekers who stole tesserae from the mosaic at Bybury. His conclusion stokes that craze even more by encouraging readers to look for Roman antiquities everywhere (Pointer 1713, 39).
Modern archaeology has confirmed that the Cotterstock villa site is of major importance. Modern investigation has taken the form of aerial photography (1976), an earthwork survey of terraces and slopes (1988), a field-walking survey to plot the distribution of potsherds (1989), observations of villa walls that were cut by the Enclosure ditch and subsequent ditch cleanings (1989), and a soil resistivity survey to identify underground areas of solid masonry (1992 and 1993) (Upex 2001, 64-72). Potsherds found at the site range in date from the first through the late fourth centuries, with the bulk (forty-two percent) dating from the late third and early fourth century.
Architectural analysis based on these various studies reveals a very large villa with ranges of buildings set around four courtyards (Upex 2001, 82-84). “The overall length of all the ranges of buildings and courtyards is in excess of 260 m,” putting it on a scale with other large villas around Britain (86). In its immediate vicinity, the Lower Nene Valley, the Cotterstock villa is “by far the largest structure of its kind” (86-87).
Upex hypothesizes that the Cotterstock villa, based on its size and apparent wealth, might have had connections with the un-walled industrial settlement at Ashton, near Oundle in Northamptonshire, two and a half km from Cotterstock. Ashton was a metalworking center beginning in the first century, and its wealth may have profited an entrepreneurial owner at Cotterstock.
Upex also considers Cotterstock’s importance within the orbit of Durobrivae, a large walled Roman town (near modern-day Petersborough) that had been a center of industry since pre-Roman times. Durobrivae and surrounding villas were identified and studied in the nineteenth century by Artis and other archaeologists, though frustratingly little is known about most of them. Upex has mapped ten such villas (including Cotterstock) all of which exhibit “higher than normal wealth patterns” through features like mosaic floors and bath suites (Upex 2001, 89). Upex speculates that these ten sites could be linked to “the numbers of ruling council members who would have formed the local government based in Durobrivae” (89).
In their comprehensive study of Roman mosaics in Britain, Neal and Cosh have classified mosaic styles according to the stylistic traits found in geographic areas; they describe the Cotterstock mosaic as belonging to the Midlands Group. They find it to be similar in size and schema to two mosaics at a Roman villa in Stanwick, and they suggest that “conceivably the same craftsmen worked at both sites” at around AD 375 (Neal and Cosh 2002, 233, 257-259).
While late twentieth-century archaeological methods have confirmed the importance of the Roman villa at Cotterstock, this 1737 engraving of the geometric mosaic at the site demonstrates the capacity of antiquarians in the eighteenth century to recognize important sites and to gather and preserve crucial knowledge. The plate commissioned by the SAL is the only visual record that survives of this impressive feature, though Artis’s reproduction of the plate has contributed to its wide dissemination. Along with the engraving of the second mosaic at Cotterstock, discovered in 1798, we have only engravings of the two mosaics to testify to the workmanship and aesthetic quality of decorative details of the villa. Although the SAL did not add the 1798 mosaic to its visual repository, its role in asserting the value of such Roman artifacts beginning in the 1730s contributed to the recognition and visual preservation of Roman mosaics in Britain in subsequent decades.
Artis, Edmund T. 1828. The Durobrivae of Antoninus. London: J. Cumberland.
Evans, Joan. 1956. A History of the Society of Antiquaries. Oxford and London: Oxford University Press.
Gough, Richard. 1790. “A Mosaic Pavement in the Prior’s Chapel at Ely.” Archaeologia 10: 1-5.
Neal, David S., and Stephen R. Cosh. 2002. Roman Mosaics of Britain Volume I: Northern Britain incorporating the Midlands and East Anglia. London: Illuminata Publishers for the Society of Antiquaries of London.
Nichols, John, ed. 1790. Antiquities in Lincolnshire, Being the Third Volume of Bibliotheca Topographica Britannica. London: John Nichols.
Nurse, Bernard. 2007a. “Excavation of a Roman Hypocaust at Lincoln.” In Making History: Antiquaries in Britain 1707-2007, edited by Sarah McCarthy, and Bernard Nurse, 127. London: Royal Academy of Arts.
Nurse, Bernard. 2007b. “The Great Pavement (the ‘Orpheus Mosaic’) at Woodchester Roman Villa Excavation of a Roman Hypocaust at Lincoln.” IIn Making History: Antiquaries in Britain 1707-2007, edited by Sarah McCarthy, and Bernard Nurse, 130. London: Royal Academy of Arts.
Pointer, John. 1713. An Account of a Roman Pavement Lately found at Stunsfield in Oxford-Shire, Prov’d to be 1400 Years Old. Oxford: Leonard Lichfield.
Society of Antiquaries of London. 1718-. Minutes of the Society’s Proceedings.
Stukeley, William. 1887. The Family Memoirs of the Rev. William Stukeley, M.D. and the Antiquarian and other Correspondence of William Stukeley, Roger & Samuel Gale, etc. Edited by William C. Lukis. Vol. III. Durham: Andrews & Co.
Upex, Stephen G. 2001. “The Roman Villa at Cotterstock, Northamptonshire.” Britannia 32: 57-91.