II. FINDS REPORTED UNDER THE PORTABLE ANTIQUITIES SCHEME

The Portable Antiquities Scheme (PAS) was established in 1997 as an initiative to record archaeological objects found by members of the public. Following a pilot phase, it was extended to England and Wales from 2003 onwards.¹ From 2004 surveys of Roman period finds documented by the PAS have been published in Britannia. This twenty-first annual summary outlines the general character of Roman finds reported in 2023 and reports significant individual artefact discoveries recorded by Finds Liaison Officers (FLOs).²

OVERVIEW

In 2023 35,981 objects of Roman date were documented by the PAS in 16,249 object records. The former exceeds the latter since a single record may document more than one item, especially in cases of coin hoards as well as in groups of building materials or ceramics. In both respects, objects and records, these totals represent a significant increase on recent years.³ The quantities for 2023, more comparable to those before 2020, likely indicate the lessening impact of the Covid-19 pandemic on object discovery, reporting and documentation. The spatial distribution of the finds reported in 2023 resembles but is not identical to that recorded in previous years. The highest numbers of records made in 2023 occurred in central southern counties, followed by eastern England.⁴ As before in Wales and in northern and western England, the numbers of artefacts documented and of records created are much lower. The general factors determining this distribution of reported artefacts remain as previously identified, i.e. Roman-period behaviour linked to the circulation and deposition of metal objects, historic and current land-use, including agricultural practice, and the constraints on of metal-detecting, the chief mechanism of object discovery.⁵

In 2023, as in every previous year, the documented discoveries of coins and brooches significantly outnumber those of all other Roman artefacts. 12,284 coins were recorded in 11,981 records (of ‘coins’) and approximately a further 16,500 in 75 ‘coin hoards’, and ‘hoards’, a return to a more typical percentage of objects and records for an annual assessment. With the addition of the examples recorded in 2023, c. 460,000 Roman coins have now been recorded by the PAS, as quantified in these annual summaries. These quantities reflect both the preponderance of coins among objects discovered and the emphasis of the PAS on recording Roman coin assemblages in toto.⁶ The distribution of coins mirrors the distribution of all finds, the highest numbers being from the same counties as noted above and the smallest in urban areas and in northern and western England and Wales. As for brooches, c. 38,900 brooches have been documented by the Scheme since 2003, including 2,006 further examples recorded in 2023. These include brooches with date spans across the late Iron Age–Roman transition as well as those dated within the Roman period proper.⁷ The distribution of brooch finds continues to follow well-established patterns, reflecting the general distribution of PAS records. The previously noted regional disparity in the ratio of brooch to coin finds among Roman objects reported to the PAS continues to occur, being much higher than average in Wales and some adjacent English counties.⁸

ARTEFACT DESCRIPTIONS

The following entries publish highlights among the past year's finds recorded by PAS staff and the Treasure Department at the British Museum as well as their research collaborators.⁹ The objects are selected for their contribution to the study of the material and visual culture of Roman Britain, in particular where discoveries made or documented in 2023 extend understanding of object form, typological and decorative variability and/or distribution. They are arranged in approximate geographical sequence, from northern England and Wales to southwestern England. The alphanumeric reference number associated with each record identifies individual object records on the PAS website: www.finds.org.uk. Some entries report objects also treated as Treasure cases and are therefore also designated with their Treasure number in the format of year (20XX) and reference number (TXX).¹⁰

The precious metal and copper-alloy artefacts presented in this report were made in the period between the late Iron Age and Late Antiquity and illustrate the diversity of function, style and motif documented in portable metal objects from the Roman period in Britain, broadly defined, from the first century b.c. to fourth or fifth centuries a.d. They include objects belonging to diverse functional categories, personal ornament comprising several brooches with dates spanning the Roman period, sometimes with a military connection, and three rings with figural devices, figurines representing divine beings and animals, mounts and handles for vessels and knives, furniture and harness fittings and a likely staff terminal, a dagger hilt and nail cleaner. We report in outline two larger groups, assemblages of pewter vessels from Suffolk and of personal ornaments and tools from Wiltshire.

LANCASHIRE

(1) Near Kirkby Lonsdale (LANCUM-F0DF1B) (fig. 1)¹¹ A copper-alloy brooch with a grey-green patina, complete apart from the pin and well preserved, 65 mm long and 14 mm at its widest, weighting 18.5 g. It is formed by two arms linked in an S-shape. Each arm extends from an outer bar-like terminal to meet at the brooch's centre where the junction forms a raised ridge, the arms’ trumpet-like flaring accentuated by curvilinear moulding where they meet. On the upper side the brooch is gently curving in profile while the underside is recessed. The fixing mechanism is lost. The centre and terminals also carry a beaded groove.

FIG. 1. Near Kirkby Lonsdale, brooch (No. 1). Scale 2:3 (Photo: P. Clarke; © Portable Antiquities Scheme)

This is an example of a brooch identified as a variant of the Dragonesque tradition, category C in Hunter's recent classification, i.e. lacking clear zoomorphism.¹² In this case its form most closely resembles that of category C3, known in a handful of examples from northern Britain, including two very similar brooches documented by the PAS, one from East Yorkshire and the other from Cheshire, and a poorly preserved example from Arnside, Cumbria.¹³ One brooch of this type from Carlisle is stratigraphically dated to a.d. 120–140.¹⁴ These brooches are greatly outnumbered by other (zoomorphic) dragonesque types, more than 350 now having been documented by the PAS, a significant increase over the numbers previously assessed by Hunter and Mackreth (respectively 103 and 60+).¹⁵ Their distribution as a whole extends from southern Scotland to Cornwall, but the main emphasis of PAS finds clusters tightly in north-eastern England from Nottinghamshire and Lincolnshire northwards to county Durham.¹⁶ Within this zone new data reveal even more strongly the concentrations in rural east and west Yorkshire and the vale of York.¹⁷ By contrast, the findspots of this S-shaped variant are dispersed on both sides of the Pennines.¹⁸

EAST RIDING OF YORKSHIRE

(2) South Cave (FAKL-0C66BB) (fig. 2)¹⁹ A figurine in female form, heavily worn, especially over its upper body, with an uneven green-brown patina, 77.7 mm high. The figure, identifiable as Spes, i.e. the personification of Hope, extends its right arm forward from the body, with hand lost but (on analogy with other figures) once holding a flower. The left arm hitches up the edge of the chiton (tunic) to enable movement, the left leg pressed forward against fabric, the right leg further back, the feet lost. The pleated chiton is visible from the waist downwards beneath a himation (mantle) draped over the upper body and arms. The himation's fabric ends in distinctive zigzag folds, for example where suspended from the right arm and in the diagonal band across the body. The head and face are delicately modelled but very worn, turned a little to right as part of the forward movement of the figure. One worn row of curls frames the forehead, with elements of a diadem visible above. Traces of the rope-like tresses which extended onto the upper body are visible on the sides of the head. At the back of the head hangs a heavy plait, gathered into bands.

FIG. 2. South Cave, Spes figurine (No. 2). Scale 2:3 (Photo: K. Leahy; © Portable Antiquities Scheme)

Surviving portrait features and pose allow the figurine's identification as Spes, otherwise documented in a variety of media but above all on coins from Augustus onwards.²⁰ This figurine is similar to, if less fine in its details than, the very small sample of other copper-alloy examples, for instance larger statuettes from the British Museum, probably from Trevenzuolo, Verona province, from the Bibliothèque Nationale de France and from the Kluge collection.²¹ The pose of the South Cave figurine is slightly less rigid than other examples, with the arms extended further from the body, conveying a slightly stronger sense of the goddess in motion. The archaising character of the figure, expressed in its frontal pose, thickly pleated dress and heavy coiffure, reflects the perceived antiquity of the goddess's cult. The figurines copy the Archaic Greek statue dedicated by Germanicus in the best known of the temples to the goddess in the Forum Holitorium in Rome in a.d. 17. One of Ovid's exile verses reveals the plights in which an individual might seek help from Spes, the chained slave, the shipwrecked mariner, the ailing, incarcerated, condemned and suicidal.²² From the reign of Augustus, however, Spes became a virtue closely linked to the emperor in the form of Spes Augusta, appearing widely thereafter on coinage and occasionally on inscriptions, the latter mainly from central Italy.²³ However, the figure was known in Britain, Spes sometimes featuring on gems as well as on the reverse images of coins minted by Carausius and Allectus.²⁴

(3) Bainton (YORYM-EBD648) (fig. 3)²⁵ A complete copper-alloy mount in the form of the bust of a youthful male, modelled more or less in the round, wearing a Phrygian cap, slightly worn with a mid-green patina. The mount is 43.1 mm tall and weighs 39.4 g. The Phrygian cap has a high narrow peak which falls forward, covering a distinctive and elaborate coiffure. From beneath the cap the hair swells in exuberant clumps, each grooved to delineate the locks which form them, with further strands beneath, pulled upwards (in the centre), and sideways. Facial features are expressively rendered. Large lozenge-shaped eyes, with recessed pupils, stand out from their orbits beneath prominent brows, set above a broad, slightly bulbous nose and a narrow mouth. The youth wears a double-layered garment with distinctive corrugated mouldings representing thick folds of cloth, hanging vertically from the shoulders and lying horizontally below the neck. The cable-like decoration on the most prominent vertical folds may indicate a difference of colour or texture. A pendant hangs from the neck. From both shoulders are apparent comma-like projections, not fully preserved and of uncertain function, on the right joining the hair to form a loop. The back is flat and plain, with some traces of possible solder.

FIG. 3. Bainton, mount (No. 3). Scale 1:1 (Photo: A. Rivett; © York Museums Trust)

Numerous hooded figures have been documented by the PAS, often challenging specific identification because of the lack of definitive attributes, especially when reduced to the form of a bust or head alone. Although Attis is often met in small-scale representations in bronze and clay, confidently identified occurrences of the demi-god from Britain are vanishingly rare, one possible example having been documented by the PAS.²⁶ Mounts and harness fittings of figures with hoods or Phrygian caps are quite numerous but they have a plethora of potential identifications, this cap being a mark of ‘eastern-ness’ associated with other mythological and divine figures, including Perseus, Ganymede and Mithras.²⁷ Since this figure is heavily draped rather than heroically nude, Attis is perhaps a more likely candidate for identification, though clothing varies considerably among images of him. The energy of the head, manifested in the expressive forcefulness, is harder to parallel among bronze examples of Attis; the luxuriant curls and thickly folded drapery may offer better evidence for this image being in the spirit of portraits of him.²⁸ The distinctive pendant (bulla) worn in this image is paralleled in one bronze figurine of Attis from Lyon but does not itself confirm the identification.²⁹

(4) Holme upon Spalding Moor (SWYOR-16DE9F) (fig. 4)³⁰ A copper-alloy zoomorphic brooch in the form of a standing quadruped, a horse or perhaps a camel, partially preserved, but lacking most of the front legs, catchplate and pin. The surface is worn, with a dark-green patina. The body is 57.4 mm long and weighs 28.1 g. The animal stands upright, leaning forward. A long, thin neck carries the head, ears pricked up, and with some detail, including eyes and mouth, partly preserved, the latter slightly open. The back carries a saddle cloth or rug, demarcated by a cross-hatched panel and with a diagonal grid within containing recessed dots. The rump slopes to a short tail and slender back legs, thickening towards the feet. There they join the tubular housing sheathing the axis bar, visible at both ends, with a likely slot at the centre to allow for articulation with the pin.

FIG. 4. Holme upon Spalding Moor, brooch (No. 4). Scale 2:3 (Photo: I. Whitehead; © West Yorkshire Archaeology Advisory Service)

The animal is difficult to identify to species. Similar brooches are mainly known through online sales, but a handful are documented through metal-detecting and excavation, including examples from Leasingham, Lincs, and Pontefract, W. Yorks, as well as finds from northern Italy and Croatia, respectively from late Republican or Augustan and second-century contexts. With their distinctive arrangement for the pin mechanism, these are a possible imported brooch type.³¹ The other examples are certainly equine but in this case the distinctive modelling of the head, the thin, elongated neck and the saddle cloth prompt an alternative identification as a camel, though it lacks the pack commonly seen in camels depicted in Roman art.³²

DONCASTER

(5) Burghwallis (SWYOR-FEF388) (fig. 5)³³ An incomplete copper-alloy plate brooch, 29.3 mm long, 24.5 mm wide, and weighing 8.9 g, with a mid-green patina. The brooch is circular, but partly in openwork with two substantial voids, enclosed by thin wing-like arcs which frame a splayed tail-like foot. This feature carries parallel incised lines on both sides. In the centre the brooch has a recess filled with decayed red enamel, in which what is left of a central copper-alloy rivet may indicate the attachment of another element. At the ‘head’ of the brooch is a projecting loop. On the back of the brooch, by the looped projection, survives part of a lug, bent over at the top to house the axis bar with a slot for the pin, both now lost. An incomplete catchplate is set behind the foot.

FIG. 5. Burghwallis, brooch (No. 5). Scale 1:1 (Photo: I. Whitehead; © West Yorkshire Archaeology Advisory Service)

A near-identical brooch (SWYOR-8B6C81) was found in November 2020, c. 400 m from the findspot of this example, clearly connected but too far away (from a different field) plausibly to have paired in a burial. Otherwise no close parallels have been found for these brooches. Michel Feugère (pers. comm., May 2024) suggests that the form might combine elements of two brooch types of mid-first-century date, perhaps as ‘an Insular creation’, geometric plate brooches, for example those in crescent form, often with a central recess carrying decoration, and the splayed foot of the Rosette or Thistle brooch.³⁴

(6) Barnby Dun with Kirk Sandall (SWYOR-8266DB) (fig. 6)³⁵ A well-preserved copper-alloy brooch carrying a solid moulding in the form of a female bust, a maenad perhaps or Ariadne. The brooch is 27.4 mm long and weighs 17.4 g. It has a green patina with some patches of corrosion. The bust is finely modelled, with the head turned to the left and tilted downwards. The long oval face, tapering to a rounded chin, is rendered with delicacy, with heavy lidded eyes and a narrow mouth. The hair falls in curling locks from a centre parting, swept back from her face on both sides and with tresses tumbling onto her shoulders. Broad leaves sit on the hair on both sides and a bunch of fruit, grapes or perhaps ivy berries, is visible on the figure's right by the neck. Finely modelled folds of a tunic hang low on the bust. The reverse is flat and plain. Behind the head is a housing for an axis bar and hinged pin, partially preserved, but the catchplate is lost to damage.

FIG. 6. Barnby Dun with Kirk Sandall, brooch (No. 6). Scale 1:1 (Photo: I. Whitehead; © West Yorkshire Archaeology Advisory Service)

The leaves and fruit, whether ivy or grapes, identify the figure as a probable maenad or perhaps Ariadne, since the latter is also sometimes represented with a garland of ivy leaves.³⁶ The slant of the head, the movement implied by the décolleté drapery, rising a little off the figure, and the somewhat dishevelled hair, may hint at the semi-ecstatic state of maenads in other media, for example sarcophagi.³⁷ The same disposition characterises other Maenads on objects, for instance on steelyard weights from Nursling, Hants.³⁸ Heads of this type can be widely paralleled on portable artefacts. On cameos and gems, for example, it is difficult to differentiate between maenads and Ariadne.³⁹ Vessel mounts sometimes carry similarly ambiguous figures, for example from Littlecote, finely rendered and potentially a maenad, satyr, or Bacchus himself, or from Birrens, another satyr or maenad.⁴⁰ While other brooches in the form of female heads are occasionally documented, none are modelled with the sophistication of the Barnby Dun piece (perhaps remodelled from a mount: H. Eckardt, pers. comm., June 2024), which emulates the qualities of portrait gems.⁴¹

POWYS

(7) Old Radnor (LVPL-45C84D) (fig. 7)⁴² A complete copper-alloy zoomorphic bucket mount, 40 mm long, quite worn with a smooth dark-green patina. It weighs 37 g. The fitting takes the form of a hollow stylised bovine head modelled in the round. The head tapers from a broad cranium to a narrow rounded muzzle. A pair of short curling horns projects forward, linked by a narrow band from which short vertical incisions render hair above the eyes. The recessed eye sockets have likely lost a setting. A broad flange extends around the face. A rivet survives on the animal's right, once attaching the mount to a vessel, but on the other side this has broken away.

FIG. 7. Old Radnor, mount (No. 7). Scale 3:4 (Photo: H. Beeton; © National Museums Liverpool)

This example joins many other oxhead vessel mounts documented by the PAS.⁴³ While sharing the generic qualities of similar mounts, this particular form is hard to parallel in its details, in particular the elongated head and the evidence for rivetted attachment. It is likely to date to the late Iron Age or early Roman period (c. 100 b.c.–a.d. 100).

SHROPSHIRE

(8) Wem (WMID-1673AD) (fig. 8)⁴⁴ A well-preserved copper-alloy two-link or ‘single-jointed’ bridle bit of the Polden Hill sub-type, with a mid- to dark-green patina, somewhat abraded. The bit comprises both the rein rings, i.e. the cheek pieces, and two links for the mouthpieces. With an external diameter of 69.7 mm the rein rings are of simple form, circular in cross-section. They are threaded through narrow terminals at one end of each link, these carrying distinctive ear-like mouldings typical of the sub-type. At the other end of the link, connected by a short-ridged shaft, is a circular collared head which interlocks with the terminal on the corresponding link, set at right angles to it. The links are of similar size, one being 72.7 mm long. The overall weight of the bit is 270 g.

FIG. 8. Wem, bridle bit (No. 8). Scale 1:2 (Photo: T. Gilmore; © Birmingham Museums Trust)

Many bridle bits have been documented by the PAS, both as single bits and in larger numbers from likely structured deposits, testimony to the ubiquity of sophisticated horse harness in this period.⁴⁵ This is the first complete example of a two-link bridle bit to have been recorded, a less common type from Britain and Ireland. Incomplete examples also probably from two-link bits have, however, been documented in several other cases.⁴⁶ The sub-type to which this bit belongs is dominated by the 16 examples from the Polden Hills hoard, Somerset, dated to the first century a.d., spanning the conquest period.⁴⁷

(9) Weston-under-Redwood (DUR-2F3489) (fig. 9)⁴⁸ An incomplete copper-alloy jug handle, broken into two non-joining pieces with figural elements at both ends. The upper part of the handle is 57.8 mm long, 65.5 mm wide where it meets the vessel rim, and weighs 112.2 g. The lower part of the handle is 54.2. mm long, 32.1 mm at its widest and weighs 43.1 g. Although the fragments do not directly connect, their similar dimensions and patina and discovery only metres apart suggest they are part of the same handle.

FIG. 9. Weston under Redwood, vessel handle (No. 9). Scale 2:3 (Photo: S.Gibson; © Portable Antiquities Scheme)

The upper part of the handle rose above the vessel, widening and spreading on either side into a crossbar where it met the rim, the underside grooved as a slot for attachment. On the crossbar the handle ends in a button-like terminal to the left and is broken on the right. At the junction with the rim the handle carries a worn head, elongated at the back, tapering as the handle narrows, divided by a central groove. At the front are traces of what may be a muzzle and other facial features. Below this head grooves define a triangular zone, its apex at the edge of the handle. On either side parallel sinuous incisions extend from the face to the crossbar.

The handle proper has a strongly moulded profile with a marked central rib on the outer surface, with volutes on either side where it broadens to form the escutcheon. The latter is modelled as a youthful head, its physiognomy and hair quite finely rendered. Long locks frame the face, hanging in a low fringe above the temples and to the sides behind the ears, all separately and fully modelled. A band or diadem separates these from the hair on the top of the head, rendered as much tighter curls by shallow incised lines. The eyes are defined by heavy lids and a recess for the pupils, with a wide nose set between slightly puffed cheeks, a full lower lip and rounded chin. The escutcheon's lower edge is worn and damaged. Silvery traces on the reverse of the mount may be remains of solder which once attached the handle to the body of the vessel; traces were also observed on the other fragment.

The handle was once attached to a copper-alloy trefoil jug of a type widely found in the Mediterranean and the north-western provinces.⁴⁹ Recently documented complete examples from Britain include the squat jug from a Claudio-Neronian burial excavated during the A2 widening scheme (Kent) and the taller vessel documented from the early third-century a.d. burial at Kelshall, Herts.⁵⁰ Fragments of handles of this and other jug types have also been documented by the PAS. Anthropomorphic terminals for the body attachment are more common, some distinctive like the bearded male face-satyr from Miserden, or the idealised youth from Moulsoe.⁵¹ Others are more generically smooth-cheeked and youthful.⁵² The most striking image remains the handle with lion head, snake back and claw from Appleby, North Lincs.⁵³ For both its anthropomorphic motifs, the Weston handle has a generic rather than specific affinity with other examples. Lions are the commonest motif where the handle meets the rim, the most likely identification here. Among the closer parallels is a better-preserved face from Vertault/Vertillum (Cote d'Or), labelled as a lion, but mooted to have an alternative identification as a bearded human.⁵⁴

HEREFORDSHIRE

(10) Whitchurch (NMGW-8937BB) (fig. 10)⁵⁵ A copper-alloy folding knife handle in the form of an erotic scene, 47.6 mm long and weighing 28.4 g. The blade's corroded remains are visible in the slot which once housed it, as well as possible traces of a rivet around which it pivoted. Despite the loss of surface detail to wear and corrosion, enough of the handle survives to identify it as one of a group showing a three-person scene of sex (or simulated sex), modelled in the round and known mainly from Britain and north-east Gaul. Better preserved examples allow the scene to be described; in summary a standing male, engaged in real or simulated sexual intercourse with a female figure, the latter using a kneeling male as a support.⁵⁶ The legs of the female figure are supported by the standing male's arms. The poor modelling and wear make it impossible to establish the character of genital contact, for example, or whether the standing figure, legs slightly bent, is bearded and cloaked as in other examples. The female figure's arms are braced against the shoulders of the kneeling figure, perhaps hooded, who looks out from the encounter with his hands clasped over his knees. His phallus is visible but it is impossible to tell if he is masturbating, as is sometimes the case.

FIG. 10. Whitchurch, knife handle (No. 10). Scale 3:4 (Photo: G. Whatley; © Portable Antiquities Scheme)

This is one of over 20 examples of likely mid-imperial date now documented, including 11 from Britain, of which at least five have been recorded by the PAS.⁵⁷ A further well-preserved example, with much of the blade intact, was documented at Tickenham, North Somerset (GLO-481969). The Whitchurch and Tickenham handles extend the distribution of such handles further west. The poor survival makes it difficult to attribute the Whitchurch knife to one of the groupings into which some handles can be divided by details of pose and dress. However, it lacks the high headgear worn by examples from Tickenham, Lambourn, and the Bathurst collection, likely from Cirencester.⁵⁸ The scene on these handles is particular to the north-west provinces. They were perhaps made as souvenirs of clowning and titillating scenes seen in spectacles in Britain and Gaul.⁵⁹

LINCOLNSHIRE

(11) Heckington (LIN-7FC85C) (fig. 11)⁶⁰ A decorative copper-alloy dagger hilt of Iron Age or Roman date, comprising the moulded pommel, handle and guard of the weapon, 47 mm long and weighing 33.8 g. Iron staining on the top of the pommel and along the guard indicates that remains of the blade tang may survive in situ. The guard and pommel form opposed peltas with the curves accentuated by ridges and incised lines. Both guard and pommel are similar, with central opposed V-shaped incised designs flanked by inward curving lobes. The handle that separates them is moulded into a mirrored series of collar, trough and collar upon either side of a bulbous central collar.

FIG. 11. Heckington, object (No. 11). Scale 3:4 (Photo: R. Trevaskus; © Portable Antiquities Scheme)

While the object resembles sword handles, its small size is more appropriate to a dirk or dagger than a longer sword blade. Later Iron Age (c. 200–100 b.c.) anthropomorphic sword and dagger hilts offer one point of comparison, though the design utilised on this object is more abstract. It most closely resembles the hilt furniture of the first-century a.d. Hod Hill sword now in the British Museum, a type identified by Miks as a Roman form with a Pompeii-type blade. Other examples of hilt furniture have been recorded with the PAS.⁶¹ The plasticity of the moulding is also found on other objects, suggesting a later date for this dagger, perhaps of the first century a.d. in line with the Hod Hill sword.⁶²

(12) Nettleham (LIN-F3A177 2023 T799) (fig. 12)⁶³ A well-preserved gold finger ring dating from the first–second century a.d., weighing 3.7 g. The band, near oval in cross-section, has been slightly squashed and flattened. It widens to form triangular shoulders, framing an oval concave bezel which carries a relief portrait of a bearded male head in profile facing right, finely modelled. The head is smooth and bald, save for locks falling behind the ear, curling upwards as they reach the neck. The beard is separated into individual full strands, the small eyes are rendered deep-set beneath heavy brows, with a bulbous nose and quite full lips.

FIG. 12. Nettleham, ring (No. 12). Scale 3:2 (Photo: R. Trevaskus; © Portable Antiquities Scheme)

The subject can be identified either as Socrates or Silenus, more likely the former, an ambiguity seen in other examples and other media.⁶⁴ It resembles, for example, portraits of Socrates in profile engraved on an onyx cameo and red jasper in the British Museum collection, respectively of first–second- and first–third-century a.d. dates.⁶⁵ Serving to communicated the bearer's paideia (intellectual sophistication), a small number of gem-cut and glass philosopher portraits survive from Roman Britain, one for example from Gloucestershire (red jasper), probably of second-century date, and another said to be from South Shields but without definitive provenance, as well as an excavated glass intaglio from Maryport, likely showing Zeno of Kition, and most striking of all, the garnet head of Socrates from Brampton, Norfolk.⁶⁶

(13) Revesby area (DUR-729817 / 2023T841) (fig. 13)⁶⁷ A complete gold finger ring, containing a superbly cut red jasper intaglio, with a maximum internal diameter of c. 21 mm, weighing 6.2 g. Ring and intaglio are very well preserved, albeit with two fine cracks across the latter. The plain oval hoop carries flattened shoulders and a wide oval bezel. According to Martin Henig (pers. comm., February 2024) the ring's form ‘is a chunky development from the later second century a.d. Henig type V and VI rings, dating most probably to the Severan period’.⁶⁸ The intaglio figures two engraved heads. A mask of a bearded male faces forward, with round thick-rimmed eyes, large and empty beneath heavy brows, a broad flat nose, and full straggling moustache and beard framing a wide mouth. This mask is conjoined with a youthful smooth-cheeked androgynous head in profile facing right, eyes lightly framed by lid and brow, and with full lips. The faces share the same stylised head of hair, combed forward into rolled fringes.

FIG. 13. Revesby area, ring (No. 13). Scale 3:2 (Photo: B. Westwood; © Durham County Council)

This combination echoes other pairings of comic theatrical masks in gems and in other media. Among the former are similar sard gems in the British Museum, one combining a frontal bearded face with a youthful head in profile, the latter identified (disputably) as female and the former as a satyr, and the same pairing with a third slave mask instead of the head of hair..⁶⁹ Similar conjoinings are illustrated on marble reliefs, pairing ‘old man’ and ‘young man’ masks at Pompeii, and a ‘slave’ and ‘young man’ on an example unknown provenance from the British Museum.⁷⁰ The masks also recall those of Bacchic figures on gems, where Silenus joins young satyrs or maenads. In its striking appearance the satyr-like old man on the Revesby ring is comparable to the subject of an onyx cameo of first-century date now in the Rijksmuseum van Oudheden, Leiden.⁷¹ The frontal face also recalls the staring mask held up in a painted ‘initiation’ scene from the Villa of the Mysteries at Pompeii.⁷² Combinations with Silenus heads in profile on other rings include examples with masks of satyrs, maenads and Pans, also cut on red jasper.⁷³ Linking Silenus masks with other Bacchic and animal figures echoes the hybrid creatures in the pairings on gems (grylloi) thought to have an apotropaic effect against Invidia.⁷⁴ Martin Henig (pers. comm., February 2024)) notes the finely cut red jaspers excavated from the drain of a Severan-period bath-house by the fort at Stanwix, perhaps carved in Carlisle itself in connection with the presence of the Severan house on campaign (a.d. 208–11). He suggests that the creation of the Revesby gem too was a product of artistic output stimulated by imperial patronage.⁷⁵

NOTTINGHAMSHIRE

(14) Misson (SWYOR-3E8C66) (fig. 14)⁷⁶ A thin copper-alloy knife handle for a fixed-blade knife, taking the form of an asparagus spear with a dark-green patina. It is 81.5 mm long, sub-circular in section with a maximum diameter of 10.9 mm and weighs 33.5 g. It carries scales on most of its length which only have started to expand into foliage at the tip, with the tiny delicate fronds modelled with great care. It is incomplete at one end where the iron blade of the knife would have joined the handle.

FIG. 14. Misson, knife handle (No. 14). Scale 1:1 (Photo: I. Whitehead; © West Yorkshire Archaeology Advisory Service)

Handles for fixed-blade knives tend to be less elaborate than those for folding knives. This is the first instance of a knife of this type found in Britain. Among the other 18 examples, those with a known findspot cluster in the Moselle and Rhineland regions, especially around Trier and perhaps also Cologne; at Trier there is also epigraphic evidence for asparagus consumption.⁷⁷ Beyond this cluster the only distant documented findspots are those at Misson and at Silistra (Bulgaria) on the Danube.⁷⁸ The handle from the Borg villa is likely to derive from a third-century context but direct dating information for the group is otherwise lacking.⁷⁹ Observation and experimentation show that moulds were formed directly from asparagus spears themselves. The latter were encased in carefully tempered clay to form the mould which was then fired to reduce the spears to ash before pouring in the metal.⁸⁰ In the details of its form, the Notts example differs from others, being longer, thinner and less ‘leafy’ with more limited ‘ferning out’ (i.e. growing foliage). In this respect it resembles the example from Rheinzabern, but is more elongated, perhaps using a younger spear as its model.⁸¹ The technical virtuosity embodied in the knife handle likely made it a symbol of wealth in its own right, alongside its allusion to consumption of a status-defining food. As a likely import its discovery need not indicate asparagus cultivation in Roman Britain, though asparagus beds were provisionally identified in excavations near Cambridge in 2014.⁸²

LEICESTERSHIRE

(15) Market Bosworth (LEIC-77F0FB) (fig. 15)⁸³ An incomplete, copper-alloy proto-P-shaped brooch, 64.3 mm long with a dark-green patina and surface traces of tinning. The brooch is sprung and missing only its pin which was recovered alongside it. The head features a projecting waisted knob, the first in a sequence of five that run down the centre of the bow arch. These knobs are joined along a base plate that is riveted to the bow. On either side are narrow strips of corded decoration that begin at the head and terminate just above the foot in a similar transverse band of cording. A flange separates the bow from the lower bow and faceted foot, which has a small cross-moulding and part of the catchplate behind.

FIG. 15. Market Bosworth, brooch (No. 15). Scale 2:3 (Photo: V. Sazfara; © Derby Museums Trust)

The brooch belongs to a type identified at Richborough (T185), dated to the early third century a.d.⁸⁴ A close parallel can be found in other brooches documented by the PAS.⁸⁵ According to Mackreth, this form is distributed primarily along Hadrian's Wall and through eastern Britain. Its western Leicestershire find spot marks out the Market Bosworth brooch as one of the more westerly examples: Mackreth lists only one from Gloucestershire and Wales respectively, compared to 111 from Northumberland and 11 from both Norfolk and Lincolnshire.⁸⁶ Bayley and Butcher suggest a military connection due to the type being solely found on military sites.⁸⁷ However, the examples recorded with the PAS do not come from known military sites and are quite widely distributed, though they may of course reflect the activities of the military within the wider province.

NORTHAMPTONSHIRE

(16) Hinton-in-the-Hedges (NARC-9D96B1) (fig. 16)⁸⁸ A copper-alloy ‘Ringfibel’ or annular brooch, complete with frame and pin, 64 mm wide, the circular frame 44.4 mm in diameter, weighing 13.6 g, with a smooth brown patina.⁸⁹ On the underside the frame is plain and bevelled, on the upper side bevelled too, with punched circles evenly spaced around its circumference. The pin is made of a thin narrow sheet of metal, rolled to form one tapering end and curved at the other to loop around the frame, forming a grooved sleeve around it. The pin's point fits within a notch on the frame edge, flanked by projecting plinths which each carry a quincunx arrangement of a larger central and smaller outer circles. From either side of the notch extend two symmetrical zoomorphic terminals. These comprise both stylised beak- or crest-like elements on necks which turn away from the hoop and towards one another and small heads which project back towards the hoop, with mouths slightly agape. Pairs of punched circles form the ‘eyes’ for these heads. Along each neck run parallel curved ridges which carry incised corded decoration. These meet curving channels which run to the tip of the ‘beaks’ or ‘crests’, accentuating their curving form.

FIG. 16. Hinton in the Hedges, annular brooch (No. 16). Scale 3:4 (Photo: E. Cox; © Northamptonshire County Council)

The object belongs to a category of late third- to fifth-century artefacts serving as brooches or belt-buckles, widely documented in frontier settings across central Europe.⁹⁰ A very small number have been documented from Britain.⁹¹ In this case the distinctive zoomorphic decoration finds its closest parallels in a group dated to the fourth and fifth centuries a.d., mainly but not exclusively documented in coastal Frisia.⁹² The close resemblance of the form and decoration of the Hinton-in-the-Hedges brooch both with these and with an example from Laon provide further evidence for late antique connections across the North Sea, joining other belt fittings with related affinities.⁹³

CAMBRIDGESHIRE

(17) Little Wilbraham (SF-58389D) (fig. 17)⁹⁴ A complete polychrome copper-alloy terret 76.1 mm long, weighing 83 g, with loss to wear and corrosion, including much of the enamel decoration. The terret takes a crescent form, each face carrying curvilinear decoration of the same form within a recessed field with a thinner outer and thicker inner border. The decoration comprises sinuous motifs on reserved metal within a red enamel field, symmetrically arranged around the short axis of the terret, more expansive in the centre, more compressed towards the tips. Within the raised motifs are six circular recesses containing yellow enamel. The tips of the crescent expand to trumpet-like terminals to meet the notched bar which attaches the terret to the vehicle yoke.

FIG. 17. Little Wilbraham, terret (No. 17). Scale 1:2 (Photo: A. Booth; © Suffolk County Council)

The terret belongs to the flat ring terrets of Group VIII, sub-group A, as defined by Spratling, or the ‘crescentic’ form defined by Lewis, with symmetrically arranged ornament, common across south-east Britain, especially east Anglia.⁹⁵ The limiting of the colours to red and yellow enamel may date this terret to the early first century a.d.⁹⁶ It joins many examples of the same form, typically combining curvilinear and enamelled decoration, that have been documented by the PAS since the completion of Lewis's doctoral study and provide further testament to late Iron Age and early Roman investment in decorated horse harness of this kind.⁹⁷

SUFFOLK

(18) Near Euston (SF-86DD2F) (fig. 18)⁹⁸ Thirteen pewter vessels, exposed during a detecting rally, found during subsequent excavation to have been deposited in a small pit.⁹⁹ Plough damage and corrosion made it difficult fully to excavate and separate vessels or to establish definitively how they had been deposited or identify use-wear on them. The hoard comprises a stack of plates and platters of varying diameters (from 250 to 510 mm as preserved) with four smaller bowls, two dishes and a cup. Within the group are several vessel pairs, including two very large platters, two bowls with octagonal rims and two dishes. Geometric decoration was noted on some vessels, including circumferential lines formed by incision, pellets and punchmarks as well as central motifs, one a complex arrangement of punched curved and straight lines that is hard to identify as a figural or geometric motif.

FIG. 18. Near Euston, vessels (No. 18). (Photo: A. Booth; © Suffolk County Council)

This is one of several pewter vessel groups documented by the PAS.¹⁰⁰ Its findspot lies within the eastern cluster of pewter hoards in southern Britain, i.e. within a zone extending across Cambridgeshire and adjacent areas of Norfolk and Suffolk, dated primarily from the mid-4th century a.d. onwards.¹⁰¹ Many other deposits of pewter vessels have been found in low-lying areas immediately to the west between Ely and Icklingham, many in wet contexts.¹⁰² Deposits of pewter vessels are documented in both ‘on’ and off-site contexts, the former including not only villas but also non-villa rural sites, as demonstrated at Cambourne west of Cambridge.¹⁰³ The dominance within the Euston group of plates or platters and bowls is typical, paralleled for example at Hockwold where one of the hoards is similarly stacked.¹⁰⁴ A Christian significance, i.e. a possible liturgical use, is sometimes argued for the bowls with octagonal rims, partly on the basis of the distinctive decoration found on one example from Sutton, Isle of Ely, and from a concentration of findspots in a region with other evidence for Christianity.¹⁰⁵ In this case however, without other supporting evidence, the argument for plain bowls with octagonal rims as having a specific religious affiliation is weak.

(19) Rendlesham (SF-241792) (fig. 19)¹⁰⁶ A copper-alloy object, identified as a likely horse chamfron (a.d. 120–300), white-metal coated and depicting Hercules in repoussé, found in several fragments. The largest is c. 250 mm long and 60 mm at its widest. The other fragments suggest that the original object was much larger and may have featured multiple decorative panels. The full-length heroic nude figure of Hercules stands facing forwards, his right hand resting on his club. Two other fragments contain the figure's head and part of his left side, including the lion skin draped over his left shoulder. He also carries arrows in a quiver on his shoulder. Hercules is depicted on the central panel of a hinged, three-part chamfron from Eining in a similar pose to that on the Rendlesham fragments.¹⁰⁷ What remains of the figure's head suggests that it may have been rendered fully in the round, reminiscent of a hinged chamfron from Straubing that depicts Mars or a smaller example depicting Minerva.¹⁰⁸ The latter features a similar strap attachment loop to the Rendlesham example. The strap attachment loop at the foot of Hercules also prompts an alternative interpretation of the object as a greave. Both greaves from a pair at Straubing, featuring Hercules and Mars, include an attachment point for a strap loop at the feet of the central figure in a similar position to that of the Rendlesham example.¹⁰⁹ The depiction of the figure's head fully in the round is also paralleled on a second greave from Straubing that depicts Mars.¹¹⁰

FIG. 19. Rendlesham, assemblage (No. 19). Scale 1:3 (Photo: R. Caravello; © Suffolk County Council)

Decorated greaves of the Straubing/Eining types are usually associated with the Roman cavalry and their ‘sports’ displays, an argument also possible for chamfrons of the same type.¹¹¹ Whether it is a greave or a chamfron, the Rendlesham armour can therefore be interpreted as linked to the Roman cavalry and forms a part of a larger group of militaria from the second and third century in the region around its findspot.¹¹²

BUCKINGHAMSHIRE

(20) Aylesbury Vale (BUC-850BF6) (fig. 20)¹¹³ An incomplete worn copper-alloy terminal in avian form, perhaps a sceptre head, with a mid-green patina. The object is 48.2 mm long and weighs 6.1 g. The terminal is in the form of a bird of prey seated on a thick base with wings tightly folded and head turned to the left, with a rounded beak, and just visible eyes. In front of the lower part of the bird's body are two projecting elements, a disc and what looks like another zoomorphic form, perhaps also a bird. Below the base is a substantial faceted and ridged collar, beneath which a small part of a shaft survives, ending in a worn break. The object's closest affinities are with bird-terminals identified as possible sceptre-heads, for example from Butterfield Down, Amesbury.¹¹⁴ Other examples have also been documented by the PAS.¹¹⁵

FIG. 20. Aylesbury Vale, avian terminal (No. 20). Scale 1:1 (Photo: A. Wood; © Portable Antiquities Scheme)

OXFORDSHIRE

(21) Swinford and Widford (BUC-67E42A) (fig. 21)¹¹⁶ An incomplete copper-alloy harness strap-slide of phalera form (a.d. 100–200). The circular phalera, 32 mm in diameter and weighing 15.7 g, features a central stud in the form of a human head, facing outwards with lentoid eyes and a prominent nose. The triangular head of hair is defined by a right-curling spiral. The stud is affixed in the centre of a circular recess, through which a short cylindrical rivet passes, securing it to the phalera. This recess is surrounded by a convex ring similar to Bishop's phalera type 11a.¹¹⁷ The reverse has two opposed projections that were once connected by a horizontal bar to form the strap-slide.

FIG. 21. Swinford and Widford, phalera (No. 21). Scale 1:1 (Photo: A. Wood; © Portable Antiquities Scheme)

This phalera is paralleled in examples from Nijmegen. One features a riveted mount in the form of a human face and two other anthropomorphic mounts are perhaps from similar phalerae, though as they are disassociated it is impossible to be certain.¹¹⁸ A further complete example features a stylised lion head in the central recess.¹¹⁹ Another phalera disc of similar design and with a broken reverse strap-slide frame has a central spherical rivet.¹²⁰ Van der Veen suggests that these fittings are for belts and fall within Nicolay's Group A military equipment from the Netherlands, i.e. dated from 12 b.c.–a.d. 120.¹²¹ However, the lack of decorative elements such as niello or silver inlay, and the style of the face, might indicate an Antonine date for this example.

(22) Blewbury (OXON-4D65B3 / T2022T373) (fig. 22)¹²² An incomplete silver zoomorphic hanging bowl mount, 37 mm long and weighing 5.4 g, with part of one side and the base lost to damage and some darker patination as well as some traces of gilding. The upper part of the mount takes the form of a slim, elongated horse's head. The latter is effectively modelled in outline, a strong jawline, flaring mouth and nostrils, ears pricked up and nicked on the edge to render fur and a narrow ridge between them to render the mane. The neck forms a vertical rib which connects the head to the lower part of the mount, which is a grooved trapezoidal plate mainly preserved only on one side and tapering to the base. With a close resemblance to a copper-alloy metal-detected find from Sharnford (Leicestershire), and more generic affinities with other escutcheons of bird form belonging to hanging bowls among Bruce-Mitford's Group 2, the mount is likely to be late Roman in date. Later mounts for these bowls also sometimes take similar equine form, for example a finely decorated silver mount from Deopham, Norfolk.¹²³

FIG. 22. Blewbury, mount, (No. 22). Scale 1:1 (Photo: E. Caswell; © Portable Antiquities Scheme)

WEST BERKSHIRE

(23) Leckhampstead (BERK-B99239) (fig. 23)¹²⁴ A copper-alloy disc with an even green patina with openwork and champlevé enamelled decoration, 82.1 mm in surviving diameter, weighing 50.9 g. The decoration is arranged concentrically around a domed and stepped boss at the centre. The latter is framed by a circle of repeated trumpet-form recesses filled with a black corroded material, perhaps niello. This inner circle is in turn set within a decorative zone comprising lotus motifs, alternating as smaller recessed and larger openwork elements. The openwork motifs are framed by fan-shaped recesses containing traces of red enamel. The recessed motifs are set in crescent-like zones framed by arcs comprising repeated trumpet motifs, again filled with the same decayed material. On the outer edge are the remnants of a lace-like openwork arcade, mostly preserved as short stubs of projecting metal, of a kind found on some contemporary brooches.¹²⁵ On the reverse a scar beneath the boss indicates the likely point of attachment.

FIG. 23. Leckhampstead, disc (No. 23). Scale 3:4 (Photo: P. Smither; © Portable Antiquities Scheme)

The decoration is reminiscent of the enamelling documented on a small group of vessels of mid-imperial date from Britain (the best preserved being the Selborne cup) and Gaul, characterised by a dense field of geometric motifs in which lotuses and trumpets dominate.¹²⁶ The function of the fitting is unclear. It recalls the form of harness or baldric decoration but seems too delicate plausibly to be identified as such. A mount for furniture or similar role may be possible.¹²⁷

HAMPSHIRE

(24) Chilcomb (SUR-05E56A) (fig. 24)¹²⁸ A copper-alloy enamelled bow brooch, 33.5 mm long, weighing 19.6 g. The short tubular wings house an iron axis bar which hinges with a copper-alloy pin, a small section of which survives. The head is set at a sharp angle to the flat bow. The latter carries a single recessed field containing millefiori enamel inlay organised in five panels of distorted red, white and blue squares, each configured differently. The central panel is a blue spiral on a white ground; others are arranged in quasi-chequerboard form. Alternating red and white squares frame each panel. The bow tapers slightly to the rounded foot beneath which a small catchplate survives.

FIG. 24. Chilcomb, brooch (No. 24). Scale 1:1 (Photo: S. Maslin; © Portable Antiquities Scheme)

The brooch is of continental type documented in small numbers, with a cluster in northern Gallia Belgica, but with outliers now including southern Britain.¹²⁹ Examples of a variant form, with a trapezoidal bow, with a distribution centred on Luxemburg, have been documented by the PAS in Hampshire and Nottinghamshire.¹³⁰ These forms are dated from c. a.d. 90 to 150, but the use in this case of chequerboard millefiori enamel, more commonly exploited on plate and disc brooches, may push the date later.¹³¹ In this respect the brooch joins a larger corpus of fittings which exploit the same decorative potential.¹³²

(25) Monxton (WREX-C1A12F) (fig. 25)¹³³ A complete copper-alloy nail-cleaner of late Roman date, 86.9 mm long, 12.6 mm at maximum width and weighing 10.6 g. It comprises two connected parts, a suspension element and the nail-cleaner proper. The former, square in section, has split terminals at either end which are set at right angles to one another. One allows connections to the nail-cleaner itself, to which it is still linked by a rivet passing through the perforated extension at the top of the latter. The split at the other end of the suspension element allowed attachment to a belt or chatelaine. The nail-cleaner proper has a tripartite bar or neck, alternating straight and indented sides, and a piriform blade with rounded shoulders, tapering to a narrow bifid tip, grooved on the front. Where neck and shoulders meet are two curved motifs which echo the animal heads found on similar examples (see below), though they are less distinctively zoomorphic.

FIG. 25. Monxton, nail cleaner (No. 25). Scale 1:1 (Photo: S. White; © National Museum Wales)

The object carries extensive decoration, on every surface of the suspension element and on one surface of the nail-cleaner proper. This decoration comprises punched rings and dots, incised lines and zones of hatching, sometimes forming circles, saltires and scrolls. The arrangement of motifs either side of the bifid tip also hints at zoomorphic forms. This well-preserved nail-cleaner resembles others with similar decoration, especially the nail-cleaner from Tortworth (Glos), although these have split attachment ends in the same plane as the blade. The stratigraphic contexts and stylistic affinities with belt fittings indicate a late Roman date; its findspot and that of another example recorded by the PAS fit the southern distribution of this artefact type.¹³⁴ The suspension element sheds new light on the nail-cleaner's method of attachment.

EAST SUSSEX

(26) Streat (SUSS-07E0EF) (fig. 26)¹³⁵ Part of a small copper-alloy pan handle with a feline's head terminal, with a dark-green patina, quite worn with extensive corrosion, 80.7 mm long, up to 23.3 mm in diameter and weighing 107.2 g. The junction with the vessel is lost and what remains of the hollow cylinder tapers slightly towards the terminal, separated from it by a thick raised collar. Within the handle a grey ceramic core is visible. Much of the head's detail is lost to corrosion, but enough survives to suggest original fine modelling as a leopard. Save for the chin the head is fringed by a mane, indicated by thick locks swept back to meet the collar, best preserved on top of the head. These locks frame two flattened ears. A broad groove runs from the top of the mane to the muzzle; the eyes are set to either side beneath a prominent brow. On the front of the face grooves render the muzzle and whiskers, with a shallow slot beneath for the mouth and a narrow jaw. On the cheeks punched circles reproduce the animal's spotted hide.¹³⁶ Terminals in feline or canine form for handled pans of this type, usually dated to the first and second centuries a.d., are much rarer than those modelled as ram's heads.¹³⁷

FIG. 26 Streat, vessel handle (No. 26). Scale 3:4 (Photo: J. Clarke; © Portable Antiquities Scheme)

WILTSHIRE

(27) Longbridge Deverill (WILT-92070A / 2022T815) (fig. 27)¹³⁸ An assemblage of copper-alloy, iron and lead objects from a possible disturbed hoard or burial, found distributed across an area of about 0.5 m. The personal ornaments, all in copper alloy, comprised two cast jointed neck-rings and two bracelets, as well as 11 bow brooches, whole and fragmentary. The distinctive neck rings, similar in size (up to c. 157.5 mm in diameter) and form but differing in decorative detail, were formed of two halves and carried distinctive beaded decoration, worn through use by a smaller adult or adolescent, to judge from their size. Although the closest parallel derives from south-west England, namely the similar-sized neck-ring from Newquay, these combine features of different regional neck-ring forms of late Iron Age and early Roman date.¹³⁹ The two small bracelets also closely resemble one another in size and form (internal widths of c. 55 mm and 63.1 mm), both being hinged with iron pins to lock their halves together and large enough only for slim wrists. Their formal and decorative affinities are with other bracelets of first- and second-century a.d. date from southwest Britain.¹⁴⁰ Among the 11 brooches were T-shaped Colchester derivatives, Polden Hill and trumpet types. The iron tools, a possible coulter and adze, were heavily corroded. The latter resembles first-century a.d. forms, for example from Hod Hill.¹⁴¹

FIG. 27. Longbridge Deverill, neck ring (No. 27). Scale 1:3 (Photo: S. Hawke; © Salisbury and South Wiltshire Museum)

The objects were covered, seemingly, by a copper-alloy dish or shallow bowl, almost 20 cm in diameter, and likely buried in a lead box-like container. Both brooches and bracelets suggest a date centred on the decades around c. a.d. 100. The somewhat incongruous pairing of ornament, some for an individual of slender physique, and heavy iron tools make it hard confidently to attribute the assemblage to a funerary or votive context.¹⁴²

BATH AND NORTH-EAST SOMERSET

(28) Bath (GLO-3F99D8 / 2023T634) (fig. 28)¹⁴³ A complete silver finger ring of the Brancaster type of late Roman date. Its internal diameter is c. 20 mm (as measured from photograph) and it weighs 6.4 g. The hoop is octagonal on the exterior and circular on the interior. The bezel, wider than the ring and standing proud of it, carries an intaglio image of a quadruped facing right within a border of punched triangles. Its head and body are modelled as a shallow central depression with stamped decoration used to render a concentration of fur on the forelimbs and neck. Incisions represent the ears in V-form and the legs, angled at the rear, ending in triangular punches for the feet. The body narrows towards the hindquarters with the tail curving back over them, beginning as incision, ending as overlapping punched triangles. The angular shoulders carry decoration in the same technique, both showing a well-modelled standing bird facing the bezel. Heads and bodies are formed by shallow recesses with short incisions delineating detail, sharp projecting beaks, plumage of folded wings and tail, and angled legs and feet, the detail being sharper in the left-hand bird.

FIG. 28. Bath, Brancaster ring (No. 28). Scale 1:1 (Photo: K. Adams; © Bristol City Council)

Many Brancaster rings, dated to a.d. 350–450, have been documented by the PAS, contributing almost half of the 54 from Britain compiled by Gerrard and Henig.¹⁴⁴ Since they completed their synthesis the PAS sample has increased significantly, including more than a dozen further examples bearing inscriptions and/or figural motifs, widely distributed across southern and central England.¹⁴⁵ The octagonal hoop, square box bezel and the decorated shoulders are variously shared by other rings.¹⁴⁶ Birds are among the commonest motifs documented on these rings, while the quadruped on the bezel joins the menagerie of other creatures within the Brancaster ring corpus, perhaps with a leonine aspect, but is hard to attribute to species. Martin Henig (pers. comm., October 2023) notes that the long tail in this case resembles that of the griffin on the Amesbury ring.¹⁴⁷ He favours an identification of the animal as an idiosyncratic representation of a sheep, and perhaps therefore to be identified as a Christian symbol, like the ‘Lamb of God’ motif on the gem in the smaller Murlough ring from Co. Down.

SOMERSET

(29) Misterton SOM-AD9455 (fig. 29)¹⁴⁸ A copper-alloy crab modelled in the round, with an even green patina, c. 65 mm wide and 55 mm long, with significant damage in antiquity. It has a humped body, squarish and smooth, with a carapace extending beyond it, toothed along its anterior edge. The mouth parts are rendered schematically by incision. The left eye-stalk partly survives but the right is lost. The better-preserved left chela (claw) expands and rises with distance from the body. Of the legs, only part of the merus survives, i.e. the closest segment to the body, those to the back upward-turned, more easily seen on the less damaged right side. To judge from better-preserved examples, the upward-turned legs likely angled sharply downwards where the merus articulated with the next segment. The overall effect would have likely been of close-set legs configured as a row of inverted Vs, suggesting an animal resting with its body on the ground, poised to scuttle. Abdominal detail is incised, with segments differentiated on either side of the central apron. However, while the figure shows observation of crab anatomy, no particular species is the source of inspiration.

FIG. 29. Misterton, figurine (No. 29). Scale 1:2 (Photo: M. Kneafsey; © Portable Antiquities Scheme)

The Misterton find is paralleled in a small number of copper-alloy crabs of varying size and better preserved. On a lamp in crab form, said to be from a Smyrna tomb, the surviving legs, careful differentiated into segments, give an impression of their likely arrangement on the Misterton crab.¹⁴⁹ Its trapezoidal carapace is truer to typical crab anatomy, while the bumpy texture of the claws evokes the ‘warty’ or ‘yellow’ crab. A similar example in the Metropolitan Museum, said to be of Hellenistic date, was a likely support for a lamp or similar furniture item.¹⁵⁰ A well-preserved crab from Marseille, hinged at the back, is so fastidious in its emulation of naturalistic detail that its mould may have been taken directly from a crab shell.¹⁵¹ From Lyon comes an applique, serving as a possible box or chest fitting, from a stratified deposit dated to the Augustan period.¹⁵² Also dated to the Augustan period, but otherwise exceptional, are the colossal bronze crabs from Alexandria, carrying a bilingual inscriptions recording the re-erection (13/12 b.c.) at the emperor's behest of the obelisk they once supported after its translation from Heliopolis.¹⁵³

Otherwise crabs find only occasional purchase in Roman art.¹⁵⁴ To the repertoire of images associated with Augustus the crab brings the symbolic force of its designation as the symbol for Cancer, the astrological house of the moon, for example accompanying portraits of Octavian on intaglios.¹⁵⁵ On the aureus minted by M. Durmius in 19 b.c., a laurelled head of Augustus is accompanied on the reverse by a crab holding a butterfly in its claws, a possible representation of the Augustan maxim festina lente (‘hasten slowly’: Suet., Aug. 25), or as a general allusion to prosperity, or even libertas.¹⁵⁶ Crabs are also a not uncommon motif for magical gems.¹⁵⁷ Otherwise their most frequent manifestation is as Cancer in carved Zodiac images, often simplified to a rounded or elliptical body with large claws and thin legs.¹⁵⁸ While Zodiacs could sometimes be created by metal appliques on a wooden frame, for example as attested in the hoard at Angleur (nr Liege), this specific function seems unlikely for the Misterton figurine, intended to be seen in the round. Finds in neighbouring provinces of objects linked to horoscopy, like the Angleur (Liège) appliques or the Grand tablets, or the use of astrological symbols in connection with time-keeping devices (e.g. the Chevroches disc, Nièvre), suggest that astrological figures could also be comfortably recognised in Britain.¹⁵⁹ Remarkably two rare Roman copper-alloy figurines linked to the Zodiac have now been documented by the PAS from Roman Somerset, the Misterton crab and the Burrington Capricorn.¹⁶⁰