Over the late nineteenth and early twentieth century several campaigns of excavation took place at Silchester, the Roman town of Calleva Atrebatum, notably those between 1864 and 1878 led by the Reverend James Joyce with the encouragement of the second Duke of Wellington,Footnote ¹⁸ and those from 1890 to 1909 led by George E. Fox and William H. St John Hope with financial support from the Society of Antiquaries of London.Footnote ¹⁹ The ambitious Fox and Hope excavations aimed to provide a complete plan of the town within the walls (fig. 3). Stone walls and evidence for substantial buildings, streets, deep pits and wells were planned, but less imposing features were often missed or not noted.Footnote ²⁰ Their annual reports in Archaeologia sometimes observed that bone and antler objects or distinctive deposits of animal bones had been found, while notes on the latter by Herbert Jones and E.T. Newton were sometimes published.Footnote ²¹

FIG. 3. Plan of Silchester, after Boon 1957.

A range of manufactured bone and antler objects and some craft waste from the Fox and Hope excavations form part of the Silchester Collection in Reading Museum and it is this material which is used here.Footnote ²² Most of the objects in the collection are typical of all Romano-British towns. Recent work on bone- and antler-working debris and finished artefact assemblages from other southern Romano-British towns points to local manufacture in a variety of contexts.Footnote ²³ The craft was not a specialist preserve and has strong links to both blacksmithing and wood-working. One face of an altar in Rome dedicated by L. Cornelius Atimetus shows two smiths at work at an anvil and furnace, while the opposite face depicts a sales cabinet full of finished items including knives with distinctive bone handles;Footnote ²⁴ at Colliton Park, Dorchester, lines of animal bones were stuck upright alongside the forge and an antler tine handle and three antler tine offcuts were among a scatter of associated non-ferrous objects.Footnote ²⁵ The same tools were used for working in both wood and skeletal materials, while bone-inlaid wooden furniture and other inlaid or veneered items provide a further link.Footnote ²⁶

BONE

Recent excavations at Silchester show that, in common with most Romano-British towns, cattle provided much of the meat consumed by the inhabitants and in consequence much of the bone used in artefact manufacture, but bones from other domestic species and wild deer antlers were also utilised.Footnote ²⁷ Across the Roman Empire systematic and efficient butchery of large domesticates was practised by specialists.Footnote ²⁸ The meat-bearing elements were removed, the skins (often with heads and feet still attached) were taken for tanning, the horns may have gone either directly or via a tanner to a horner, and the remaining bones were used in various ways, for marrow extraction, soap, broth, grease or glue; it was often after they had been deliberately broken in the course of these latter operations that they were used for making artefacts.Footnote ²⁹ Different stages of the processing from slaughter-house to butcher, to glue factory, to bone-worker probably took place at separate locations, with any residue also disposed of separately. That the raw material available outstripped the demand for bone artefacts is, however, shown by the quantities of suitable but unused animal bones from excavations.

The few comments in the nineteenth- and early twentieth-century annual reports about bone or antler artefacts, or craft debris from working skeletal materials, show that Silchester conforms to the norm of the civitas capital as a centre for carcass processing and for artefact manufacture and consumption. In his report of 1866 Joyce recorded a dump of cattle bones in a passage in Block 1 (House 1) at the north-west corner of Insula XXI which included chopped long bones and metapodials, ‘all chopped in the same way’, a description which matches dumps of bone-working waste from London and Canterbury but could equally apply to bones chopped for marrow extraction or other industrial processes.Footnote ³⁰

Of the 1891 material Jones remarked: ‘Many of the bos bones show marks of the knife and the splint bones have often been detached for the manufacture of pins and other small objects. One such was partially trimmed down for use.’Footnote ³¹ Many of the knife marks are likely to be from butchery, but what Jones meant by cattle splint bones is uncertain, as these bones are specific to horses.Footnote ³² The only clear fact is that a partially worked point, perhaps a pin or an awl, was found. Four such roughly crafted points are in the Silchester Collection, but none is labelled with location data (fig. 4, 1–4).

FIG. 4. Bone-working debris and unfinished or distinctive objects from Silchester. Scale 1:2. (Drawing: Joanna Bacon)

During the 1896 season many perforated sheep scapulae (fig. 4, 5) were found in the lower fill of pit KK in the south-east corner of Insula XVI between House 3 and the street.Footnote ³³ The holes were:

apparently made by cutting out rings with a centre-bit or some such instrument. The rings were of two sizes, ½in [12.7 mm] and 1⅛in [28.575 mm], and were cut with two tools, one for the lesser rings, the other capable of cutting both sized rings simultaneously. In each case a central hole ¼in [6.35 mm] was made by the cutter. In the thinner portions of bone the rings were cut straight through: but in the thicker parts, after the ring had been partly sunk, the bone was turned over and a second cutting made to meet the other.Footnote ³⁴

This debris was unusual enough to warrant 35 pieces being added to the museum collection, but, as with the antler waste, there was no evidence as to where the bone-working had taken place. A scapula from which 8-mm rings had been cut came from the first-century fill of the inner earthwork and provides a possible date for the pit KK material.Footnote ³⁵ Any ring so produced must have either been split into two thin plates in order to remove the central rough cancellous bone, or been left full thickness leaving the rough tissue exposed on the outer and inner edges. Their function is obscure. Boon thought that they may be eyelets,Footnote ³⁶ but no such items have been found at Silchester. An unstratified bone ring from the bath-house site at Cramond, Edinburgh, is perhaps a near parallel, its diameter lying between the two sizes, but it is made of solid bone.Footnote ³⁷

Direct evidence for local manufacture in the form of roughly crafted or unfinished items in the collection is scarce, but can be enhanced by some regional and some distinctive objects. Two roughly worked bone shafts are probably unfinished hairpins (fig. 4, 6–7), while two needles are only roughly shaped but have been utilised (fig. 4, 8 and an unillustrated example). Twenty-four hairpins are of a regional type found among the Atrebates, distinguished by lattice-decorated reels and often a conical top ornamented with a spiral groove (fig. 4, 9–10). Similar hairpins have been found at Silchester on Insula IX during the University of Reading's excavations and also at Winchester, where the earliest example came from an early second-century soil layer.Footnote ³⁸ An unfinished small-waisted pyxis from Silchester has external rebates at each end to allow a base and lid to be fitted on, but it has not yet been neatly hollowed out inside (fig. 4, 11). Pyxides were often made of boxwood, but there are also straight-walled, slightly waisted or tapering examples in bone and ivory.Footnote ³⁹ The Silchester piece may be evidence for a local turner working in both wood and bone. Craft waste from turning is not often found, but there is a group of fragments from Colchester.Footnote ⁴⁰ The association between wood- and bone-working is also attested by some distinctive pieces of inlay ornamented with fluid guilloche designs and other motifs, several of them curved as if from roundels framing a central design (fig. 4, 12–17).

ANTLER

Of the material excavated in 1891 on Insula I and parts of Insulae II and III, Jones highlighted the evidence for the gathering and working of shed red deer antlers.

Stag antlers, both worked and unworked, occurred almost everywhere but no place was uncovered which could be said to have been specially used as a manufactory of stags’ horn implements.Footnote ⁴¹ Many of the largest and finest antlers were found in the southern part of Insula I; … with two or three exceptions the antlers had all been shed, and not cut from the head of the animal after death, and nearly all show marks of the saw or knife. Many pieces are partly worked into knife-handles and other objects.Footnote ⁴²

Here is clear evidence for the harvesting of shed antlers from the countryside around Silchester, while the statement also implies that antler debris was found in archaeological layers, rather than in the fills of negative features.

The following season's work was concentrated on Insula IV's forum-basilica and again saw the recovery of numerous antlers, this time east of the forum ‘on a Roman surface’; they had again been shed and were ‘more or less prepared for manufacture’.Footnote ⁴³ With the 1891 excavations on Insula I lying north of the forum-basilica, and the small Insulae II and III on its west side, the recovery of more antler debris on the east side points to a remarkable concentration of craft waste in the heart of the town. A red deer skull with the antlers sawn off found in 1899 on Insula XXI north-east of the forum can be added to the group, while more shed red deer antler and sawn antler waste were found in 1902 on one of the insulae inside the East Gate.Footnote ⁴⁴ Of all this material only some rough-outs which may be unfinished handles (fig. 5, 1–4) and the tips of some tines (fig. 5, 5–7) were retained. Rough-out 2 is marked ‘Forum 40’ while 4 is labelled ‘knife handle’.

FIG. 5. Antler-working waste from central Silchester. Scale 1:2. (Drawing: Joanna Bacon)

A bone or antler workshop which had suddenly been abandoned might contain tools, raw material, blanks, offcuts and blundered or unfinished pieces, and perhaps some end products, as was the case with antler-working workshops in Dacia and Gaul.Footnote ⁴⁵ Waste debris is, however, very scarce on urban excavations despite the high numbers of bone and antler objects produced and it rarely occurs in sufficient quantity to point to commercial manufacture. Such a paucity may be accounted for by soil conditions and by workshop practices, such as dumping debris in abandoned buildings or in the suburbs (as yet little explored at Silchester), using it as hardcore, burning it as fuel or disposing of it on occasional bonfires.Footnote ⁴⁶ As a maker of simple antler objects might need only a work-bench, if that, Jones's remark that no workshop generating the antler waste from the centre of Silchester was located is hardly surprising. Even careful area excavation may not necessarily pinpoint an exact spot, while Hope and Fox's methods of digging and recording, and their selective finds retention, all militated against such a building being identified at Silchester. In addition, there is considerable evidence for craft-workers disposing of debris some distance from its source — possibly opportunistically, possibly as part of formal civic policy. For example, bone-working waste was dumped in the upper Walbrook valley in London and in the suburbs in Winchester and Canterbury, sometimes in order to raise the ground level of waterlogged areas.Footnote ⁴⁷

Although the central Silchester antler debris is unstratified, a reasonably convincing case may be built to date it to the late Roman period and, more specifically, the late fourth century. Jones observed that it ‘occurred almost everywhere’,Footnote ⁴⁸ which suggests that once the turf was removed, it was visible in the ploughsoil or in the uppermost levels of the insulae, rather than in the trenches dug to follow the lines of walls. The considerable quantity recovered can be linked to a general increase in the use of antler in the late Roman period,Footnote ⁴⁹ while the manufacture of antler-handled knives may be associated with the iron-working which took place in the forum-basilica in late Roman Period 7.Footnote ⁵⁰ Other antler items in the Silchester Collection which may be associated with the debris are two pack-needles made from tine points (fig. 6, 1–2).

FIG. 6. Antler pack-needles and handles from Silchester. Scale 1:2. (Drawing: Joanna Bacon)

Within the collection two plain antler handles retaining the tang and part of a blade are very similar to some of the pieces of waste debris (fig. 6, 3–4). Such simple handles cannot be closely dated, but five curved folding knife or razor handles made from the tips of antler tines can be independently dated to the same period as the forum-basilica iron-working (fig. 6, 5–9). One with blade intact (5) is labelled ‘Well 2 INS XXIII Silchester 1900’ and is the ‘pocket-knife’ noted in the second Silchester late fourth-century hoard of blacksmith's tools and other ironwork found in the well.Footnote ⁵¹ A local product, it provides a date for the other four, all probably made by the same hand. Similar handles come from late Roman contexts at Cirencester, Lydney, Richborough and Baldock, a fairly wide distribution which suggests conformity to a general southern British style, but the Lydney handle is such a close match for fig. 6, 5, even to the suspension ring, that it too, at least, could have been made at Silchester.Footnote ⁵² The dumps of antler debris noted by the Victorian excavators may thus represent an important element in the manufacturing and economic life of late Roman Silchester.

CONCLUSION

Set in the context of selected groups of excavated material from other southern Romano-British towns (Table 1), the bone and antler objects in Reading Museum's Silchester Collection can be seen to consist largely of ‘the usual assortment of pins and needles and small flat counters’Footnote ⁵³ and other items easy to make from bone splintered during intensive carcass processing. The high concentration of scapula waste from pit KK on Insula XVI points to a short-lived episode of manufacturing rings, probably in the same north-west quadrant of the town in the early Roman period, while the concentration of antler debris on the central Insulae I–IV implies that there was a workshop in that area, probably within the late Roman forum-basilica, while several handles represent the link between blacksmithing and antler-working.

TABLE 1 BONE AND ANTLER OBJECTS FROM MAJOR TOWNS IN SOUTHERN ROMAN BRITAIN, PRINCIPAL COMPARATIVE GROUPS ONLY

Data: Silchester, Reading Museum; Chichester, Down Reference Down1978, Reference Down1989, Down and Magilton Reference Down and Magilton1993, Down and Rule Reference Down and Rule1971, Seager Smith et al. Reference Seager Smith, Cooke, Gale, Knight, McKinley and Stevens2007; Winchester, Rees et al. Reference Rees, Crummy, Ottaway and Dunn2008; Cirencester, McWhirr Reference McWhirr1986, Viner Reference Viner and Holbrook1998; Dorchester, Woodward et al. Reference Woodward, Davies and Graham1993, Durham and Fulford Reference Durham and Fulford2014; Canterbury, Greep Reference Greep1995; London and Southwark, Manning Reference Manning1985, http://archive.museumoflondon.org.uk/laarc/catalogue (accessed 25/11/2014); Colchester, Crummy Reference Crummy1983, Reference Crummy and Crummy1992; Exeter, Bidwell Reference Bidwell1979, Holbrook and Bidwell Reference Holbrook and Bidwell1991.

Hinges, lathe-turned cladding and veneer or inlay from wooden furniture also frequently appear in urban assemblages. Although evidence of manufacture occurs less often for these more specialised objects, the unfinished pyxis from Silchester shows that lathe-turned objects were produced there. The craft debris from Silchester does not match the quantities found at Canterbury, Winchester, London and Colchester, but the range of personalia and domestic equipment available shows that the town was also a centre of demand, industry and consumption.Footnote ⁵⁴