Iron - and bronze-working dominate the metalworking evidence, but there are also traces of lead-working and pewter manufacture. There is virtually no evidence of the items produced, but there are strong indications that recycled metal played a part in production.
7.1.1 The forum-basilica complex
In Figure 7.2 the metalworking evidence is shown in relation to building type. Nine forum-basilica complexes and the fortress principia at York have produced evidence of metalworking. In none of the other towns, except possibly Verulamium, has the complex been excavated extensively enough to produce similar detailed evidence. The
FiGURE 7.1. Graph showing the occurrence of industrial activity within the public buildings of towns in Roman Britain (drawn by A. C. Rogers).
Concentration within the forum-basilica complex compared with other public buildings, however, does suggest a deliberate desire to focus the activity here.
At Caerwent (Figure 7.3), iron-working activity was found in Room 9 in the east end of the north aisle of the basilica where there was a furnace and pits containing iron-working waste; associated with this were several circular hearths, surrounded by stake-holes, within
FiGURE 7.2. Graph showing the type and number of cases of industrial activity occurring within each category of public building in the towns of Roman Britain (drawn by A. C. Rogers).
FiGURE 7.3. Plan of the excavated section of the forum—basilica complex at Caerwent showing the location of the metalworking activity in the fourth century (drawn by A. C. Rogers; adapted from unpublished excavation reports, P. Guest n. d.).
The nave (Brewer 1990: 82; P. Guest n. d.). The activity appears to have dated to between a. d. 330 and 360 and was then followed by the demolition of at least part of the building at the very end of the fourth or start of the fifth century, as indicated by coins of a. d. 389402. At London, excavations at Whittington Avenue revealed a section of the east portico of the basilica that seems to have been demolished in the mid-third century. Hearths were then constructed here that left slag from iron-working (Brigham 1992: 91; Museum of London Archive and Research Centre record: WIV88). Contemporary with this were iron-working hearths in the east range of the basilica (Brigham 1992: 91). Excavations of the forum-basilica at Wroxeter (Figure 7.4) uncovered a furnace associated with iron slag in the east of the portico, in front of East Room A of the basilica against the back wall of the colonnade. Coins suggested a date of the third quarter of the fourth century and later for this activity (Atkinson 1942: 108-9). The large collection of metalwork from Room 1 of
FiGURE 7.4. Plan of the late features associated with the forum—basilica at Wroxeter, including an iron-smithing furnace (drawn by A. C. Rogers; adapted from Atkinson 1942, plate 73 and R. White and Barker 1998, Figure 44).
The west side of the basilica was probably for recycling, and there was also a furnace on the outer wall of this room. At Lincoln, a clay and tile hearth, scrap metal, and slag were found in the excavations of the east range of the forum dating to the late third and into the fourth century (Steane 2006: 186).
Better known is the evidence for iron-working from the basilica at Silchester (Figure 7.5; Fulford and Timby 2000). Victorian and Edwardian excavations had removed much of the stratigraphy within the building; the extent of lost evidence will never be
Known. What survives consists mainly of features cut into the basilica make-up that contained iron slag. Altogether, 42 kg of forging slag, suggesting smithing rather than the smelting of iron, was recovered from features. These seemed to demonstrate a concentration towards both ends of the hall, with less activity in the centre (Fulford and Timby 2000: 74). Twelve of the features containing iron slag have produced coins dating to the fourth century, between the a. d. 320s and 370s, with one small pit or post-hole (F58) containing a coin of Theodosius (a. d. 383-8). Pits containing scrap from bronze-working appeared to be earlier in date, perhaps from the late third century (ibid.: 72).
Room 12 in the south wing of the forum-basilica at Leicester also contained bronzeworking evidence in the form of a hearth and bronze fragments. This was located above the robbing of an earlier floor and can be dated by coins to the late third century (Hebditch and Mellor 1973:18). The excavation of a small area of the basilica at Exeter produced evidence of a large pit, Pit 22, containing bronze-working waste dug down into the demolition debris of this part of the building (Bidwell 1979: 110-11). Broken tiles covered by a layer of burnt clay to the side of the pit appear to be the remains of a hearth, and it seems likely that the pit was dug to extract clay for crucibles and furnace linings. There is no good dating evidence for the metalworking, although stratigraphically, from the dating of earlier phases, a late date is likely - perhaps in the late fourth or earlier fifth century. Later burials dating to around a. d. 450 cut into the building near to these metalworking features, although this need not necessarily indicate that the activity had ceased.
Excavations of the southwest corner of the basilica building at Cirencester, Rooms 2 and 3/4, produced traces of hearths and finds of scrap bronze (Holbrook 1998: 108-9). Finds including a fragment of bronze statue in the form of an eye, found in the late nineteenth century by Cripps (1898), are suggestive of metalworking. Both excavated rooms contained coins of the House of Theodosius. This, together with the fact that the hearth within Room 3/4 seems to have been replaced by an oven, would suggest that the activity went on to a late date. Evidence for lead-working has been found at Dorchester that was probably associated with the forum. Its exact context is uncertain because little is known of this building, but a hearth was dug down into a gravelled surface and was associated with lead waste and late pottery (RCHME 1970). Other types of metalworking within the forum-basilica complexes include pewter manufacture, with moulds found in the Silchester basilica and the York principia building (Fulford and Timby 2000: 73; Phillips and Heywood 1995: 66-7). At York, the fortress principia has produced evidence for iron-working in the rear range with hearths, charcoal, and iron-working waste. This probably dated to the late fourth and fifth centuries, although there is little definite dating evidence (Phillips and Heywood 1995: 66-7); in the east range there was waste from bronze-working (ibid.).
7.1.2 Bathhouses
There is less evidence for metalworking in the bath buildings, with only three examples known (Figure 7.2), but fewer bathhouses have seen extensive excavations. In the St. Margaret’s Street bath building in Canterbury both the robbed laconicum and the piscina contained evidence for iron-working that dated to the second half of the fourth century, with iron-working waste and hearths (Figure 7.6; K. Blockley et al. 1995: 185). Within the laconicum were three phases of timber structures associated with hearths and iron slag. The dating of these structures is problematic, although a coin of a. d. 345-8 was found in
FiGURE 7.5. Plan of the Silchester basilica in the fourth and possibly fifth centuries showing traces of timber structures, metalworking, and a possible tiled shrine (drawn byA. C. Rogers; adapted from Fulford and Timby 2000, Figure 71).
Silt beneath the second phase structure and two coins of the House of Theodosius were found over the third phase structure (ibid.). It is uncertain, however, whether this layer sealed the structures or represents its use.
Iron-working was also part of the industrial activity that took place on the site of the Huggin Hill public baths in London in the late second to third centuries. This evidence was uncovered in the 1989 excavations on the site of Dominant House near Upper Thames Street (Museum of London Archive and Research Centre record: DMT88) in an area close to that excavated in 1964, as detailed by Marsden (1976). Marsden recognised no industrial activity, although it is uncertain if it was missed or if there were no traces on this site. There was also a limestone mould for pewter manufacture (Museum of London Archive and Research centre record: DMT88). It is uncertain how long metalworking took place here, but it may well have continued throughout the third century until some masonry buildings were constructed on the site in the fourth century (Rowsome 1999).
At Wroxeter, the basilica associated with the public baths has produced evidence for metalworking (Figure 7.7). Within the two rooms of the annexe of the basilica there were at least two phases of bronze-working belonging to the late fourth century (Barker et al. 1997: 72-9). There is little direct dating evidence from the annexe itself for this activity, but a coin of Gratian (a. d. 367) came from the layer within the nave believed to be contemporary in date (ibid.). The evidence within the annexe consisted of two phases of timber structures associated with hearths, casting pits, and bronze waste. From the early fifth century there is evidence of activity within the baths-basilica itself, consisting of timber structures, hearths, bronze and lead waste, casting moulds, and a lime pit that had later been adapted for lead-working (ibid.: 91-5).
7.1.3 Temples
Examples of metalworking have also come from what were probably religious buildings. Excavations at 1 Westgate Street in Gloucester were undertaken in the vicinity of what has been argued was the peribolos of a temple precinct (Hurst 1999b: 155-7). Beneath the demolition material of the structure were three oolitic limestone moulds for the casting of pewter and a group of iron objects suggestive of scrap metal (Heighway and Garrod 1980: 79). This demolition has been dated to the late fourth century because of the presence of late Roman pottery, including shell-tempered ware. The nature of the destruction means that there is little other evidence for the latest phases before demolition, although one hearth was uncovered, and so the extent of the industrial activity is unclear. The insula VI courtyard structure at Cirencester (probably a temple) had a timber structure VI.7 built within the courtyard in the fourth century associated with slag and fragments of iron as well as copper alloy (Corinium Museum Archives record: CIR 1974 K). Burnt soil and lumps of charcoal may represent the presence of hearths here.
More inferred evidence for metalworking comes from the temple at Wroxeter that was excavated in the early twentieth century. Here, a large collection of bronze and iron objects, including iron hooks, loops, clamps, rivets, around 1,300 nails, and fragments of bronze statuary, was found in a location that was described as being 'between the north wall and the pedestal in front of the podium’ (Bushe-Fox 1914: 2-9). Without evidence of metalworking waste, it is uncertain whether the activity took place here or the material was collected to be taken elsewhere, perhaps to the forum-basilica. At the Butt Road
FiGURE 7.6. Plan of the excavated parts of the St. Margaret's Street bathhouse in Canterbury with the evidence for activity within the laconicum and portico in the later fourth and fifth centuries (drawn by A. C. Rogers; adapted from K. Blockley et al. 1995, Figure 92).
Structure in Colchester a piece of lead sheet, perhaps from the roof, was found, folded and placed, probably in the fifth century, within a hole made by a removed post (N. Crummy et al. 1993: 184; P. Crummy 1997: 124). The deposition of this material within a pit, a type of context known to have received deposits (cf. Fulford 2001), may indicate that it had religious connotations, but this need not mean that it was not also intended for later metalworking.
7.1.4 Other buildings
A lead ingot was found deliberately buried in a carefully sealed pit within the late phase of a room of the Vine Street courtyard structure in Leicester (T. Higgins personal communication); this may have been ritual activity but it also suggests that there had been lead-working in the area. Further traces of metalworking have been found in the latest Roman phases of the southeast corner of this building, associated with timber structures and partitions constructed within it. Iron - and bronze-working took place here, represented by waste metal, hammerscale, and a possible hearth (T. Higgins personal communication). Also in Leicester, at the macellum there was a furnace in the external portico of the west range associated with a bloom of cupellation waste (15 percent copper, 55 percent lead, and with a trace of silver) apparently dating to the late fourth or early fifth century, which Wacher has argued came from the extraction of silver from coinage (Wacher 1995: 353). Although this is unproven, the bloom does show that some form of metalworking, perhaps recycling, was taking place here. Glass waste also indicates that the furnace was used for glass-working at some point. This may indicate the continuation of craft activities in the macella to a late date.
Excavations of the Ridingate in Canterbury, in the southeast of the town, revealed that in the late third or early fourth century the southern carriageway was blocked off, creating a room that was then used for bronze-working. Evidence consisted of ash and charcoal, a hearth base, and bronze waste (P. Blockley 1989). At the South Gate at Silchester, where there are indications that the gateway may have been blocked in the fourth or fifth century, excavations beside the gate produced waste from the working of copper alloys and lead, including lumps of metal in fourth-century deposits (Fulford 1984: 75). There were also three fragments of non-ferrous hearth bottom from lead or copper alloy working. Because no hearths were found in situ, it is uncertain whether the metalworking actually took place at the gate or whether the waste was perhaps dumped there from elsewhere (ibid.: 120).
7.2 Other industries
Unlike some other regions of the Roman Empire in the later Roman period, such as North Africa (Leone 2003), there seems to be only limited evidence for pottery and lime production within the public buildings of Britain. One lime pit has been identified within the baths-basilica at Wroxeter dating to the early fifth century, and this most likely relates to the layers of slaked lime discovered within the nearby annexe (Barker et al. 1997: 96). The duration of use was short, however, as the pit was subsequently used for casting lead. The small quantities of lime represented here are perhaps more likely to relate to timber building construction, such as whitewashing, rather than building in stone. Glass-working in the Leicester macellum has been mentioned and glass-working was also identified, together with metalworking, at the site of the Huggin Hill public baths in London (Museum of London Archive and Research Centre record: DMT88).
During what was possibly the latest phase of occupation within the monumental complex at the Winchester Palace site in London, bone-working is attested (Figure 7.8; also see Yule 2005: 76, 78-9). The stratigraphy is not entirely clear but the evidence consists of rough-cut and unfinished bone pins as well as a number of complete pins. These appear to be associated with the demolition layers of the complex, indicating that production had
FiGURE 7.7. Plan of the Annexe of the Wroxeter haihs-basilica in Phase Wi dating to the late fourth century, showing timber structures, hearths, and casting pits for hronze-working, and the location of hahy burials that seem to have been associated with the industrial activity (drawn hyA. C. Rogers; adapted from P. Barker et al. 1997, planA6i).
Taken place immediately prior to demolition or perhaps within the ruins of the building (ibid.). Bone-working is also attested in the latest phases of the portico of the St. Margaret’s Street baths in Canterbury, associated with the timber stalls and again from the area of the palaestra (K. Blockley etal. 1995: 199-201). The evidence consisted of large quantities of cattle metapodials, fragments of bones, and numerous unfinished pins. The fact that the material was fairly spread out, and that some became incorporated into later layers, suggests a degree of disturbance on the site and the possibility that the material represents rubbish derived from elsewhere. Alternatively, bone-working may indeed have taken place within the baths portico, with the waste then spread out into the nearby areas. Large-scale butchery was identified within the latest Roman layers of the fourth-century Flaxengate basilican building at Lincoln (data from the Lincoln Archaeology Services archives).
A few needles from the late phases of the baths-basilica at Wroxeter may indicate some kind of cloth-working. They came from floor surfaces, making it more possible that they related to activity within the building. There were three needles from the nave, one from the north aisle and one from the north portico belonging to Phase W (late fourth century), and one needle from the north aisle and two from the north portico belonging to Phase Y (early to mid-fifth century). Three further needles came from the Phase Z (fifth or sixth century) rebuild. Needles are amongst the finds assemblages from other sites,81 although it is uncertain whether they represent anything more than small-scale needlework, were unrelated to this activity, or simply dumped on the site from elsewhere. Spindle whorls have been found amongst finds assemblages in some public buildings, but mainly in earlier periods, and some were in make-up debris and other deposits. The Notitia Dignitatum mentions a gynaeceum (cloth factory) at Venta, procurator gynaecii (in Britannis) Ventensis (Seeck 1876: 151), which has been interpreted as Winchester (Biddle 1975: 299), but it may well have been referring to another town. No definite structural traces have been found, although speculation has led to the suggestion that a late Roman oblong structure containing hearths and ovens, apparently built on the site of a demolished temple, may have functioned as the gynaeceum (ibid.).
World History
