Discussion and Conclusions (Part 2)
 

Metalworking

Metalworking on the Walmgate site was concentrated within a complex of timber-framed buildings (particularly Buildings U and W) from the mid 14th to the late 16th century. This date range overlaps with two other metalworking sites in York, Bedern Foundry (AY 10/03) where a major bronze working industrial complex was discovered dating from the 13th to the early 16th century, and St Andrewgate (AY 10/07 forthcoming) where a sequence of industrial workshops was located which carried out both iron smithing and copper alloy casting from the early 14th to the early 16th century. Finlayson (AY 10/07 forthcoming) has suggested that there may have been an improvement in economic conditions in the late 13th or early 14th century, resulting in a greater demand for metal products. Evidence from Walmgate suggests that economic growth continued into the 15th century.

The non-ferrous metalworking evidence from Walmgate is typical of a medium-sized casting workshop. The nature of the copper alloy brought to the site is unknown, but there was clearly some scope for recycling miscast or broken items. Some smaller items could have been cast using small copper-rich 'cakes' as ingots, but larger items like cauldrons needed much more metal. Cauldrons and small items can be cast using a wider range of alloy types than are used in bell casting, which requires very careful alloy selection.

Heating was carried out in flat-bottomed crucibles, with the crucibles probably placed directly onto the floor of a furnace or hearth such as 1563 in the southern room of Building U, 1859 in Building W or 1394 in Building V. The large construction cut (2458) in the southern room of Building U may also have contained a furnace. The chimney stack (1487) between Buildings U and V, an L-shaped structure (1837) in Building V and footing (2200) in Building W may all have supported waist-high furnace structures. If so, these furnaces might have been used for heating crucibles as well as for ferrous metalworking. The crucibles were probably made close to the source of the clay rather that at the site.

Moulds, however, were probably made on site. It is not possible to determine exactly how the Walmgate moulds were made, but moulds from Prudhoe Castle were made in several sections (J. Dobie, pers. comm.). Carrying and pouring the molten metal would be the most dangerous procedure, so the moulds would be placed close to the furnace. There is no clear evidence for casting pits at Walmgate, so presumably the moulds were simply placed on a flat surface. After cooling, the mould would be broken off and the cauldron finished, by sanding, polishing and fitting a handle. The moulds could not be re-used. The exact design of the cauldrons produced is not known, but the Walmgate moulds are comparable to others of this period, particularly vessel mould fragments from excavations at Bedern (AY 10/03). Cauldron moulds and thick-walled crucibles identified at St Andrewgate were also very similar to those from Walmgate.

The amount of material recovered at Walmgate and St Andrewgate is relatively minor compared with that from Bedern Foundry, where a much larger area was excavated. Around 13,000 cauldron mould fragments, for example, were recovered at Bedern. Again, the types of mould and crucible found were very similar to those at Walmgate. The following table gives an idea of the relative amounts of mould, crucible and slag found at each of the three York sites.

Table: weights (in kg) of three key metalworking debris types at three York sites
MouldsCruciblesSmithing slag
Walmgate25.40.8138
Bedern2301.76.4
St Andrewgate2.85.994.5

Relatively little crucible debris was recovered from Walmgate. Crucibles would have been used as many times as possible, especially if they had to be brought in from an outside pottery, and the thick late medieval crucibles would have been quite robust. Even so, it seems unlikely that a crucible could have been used more than five or six times before the aggressive environment in the furnace caused damage. Damaged crucibles may have been broken up and the vitrified contents crushed and remelted in order to extract the copper alloy trapped within them.

Two other English cities, Worcester and Exeter, have produced evidence for medieval non-ferrous metalworking, in the form of both structural evidence and metalworking debris. At the Deansway site in Worcester (Taylor 1996), four tonnes of mould fragments were recovered, mostly relating to vessel casting, but with some for bells and other objects. Several furnaces were also found, which were probably used to cast large objects. 'A number of crucibles' were found, presumably a small number. The casting was mainly carried out in the 14th to 15th century. At Exeter, casting evidence comes from four sites (Blaylock 1996): the Church of St Mary Major in the cathedral close produced 12th century evidence of a casting pit for bells; at Mermaid Yard there were waste dumps of mould fragments (for bells and cauldrons) but no structural evidence; at Paul Street a bell-casting foundry dating to the 17th and early 18th century was found, but no waste dumps; at Cowick Street (early 16th to early 17th century), several casting pits for bells were found, together with two furnaces and more than 500kg of mould fragments, mostly from cauldron moulds.

The only site in York, other than Walmgate, to have produced significant quantities of iron smithing waste is St Andrewgate (AY 10/07 forthcoming). There iron smithing was carried out within a workshop during the later 14th to early 15th century and in the close vicinity of the excavation area in the late 15th to the early 16th century. At Walmgate iron smithing was carried out in Building N in the late 12th to early 13th century and in Buildings U and W from the mid/late 14th century through to the late 16th century.

At other excavated rural smithies in England (see AY 10/07 forthcoming, citing Astill 1993, for a comprehensive overview) nine categories of non-structural evidence have been identified as pointers to show the presence of smithing: smithing slag, hammerscale, ash layers, bar iron, scrap iron, blanks, incomplete forgings, metalworking tools and associated stone artefacts. Only two of the seven sites investigated produced eight of these categories, the others having between three and six. Walmgate has evidence for seven of these categories. It has also been suggested (Biek and Fells 1980) that small amounts of slag may represent large quantities of manufactured items, recovered smithing slag equalling approximately 1% of forged metal (10kg of slag = a tonne of artefacts produced). What survives therefore is not an accurate representation of the full range of ironworking which may have been taking place (Astill 1993). Recycling and re-use of materials was high since iron becomes purer the more it is worked. Hammerscale and hearth linings provide the best evidence for ironworking as they do not generally travel far from their source of origin (AY 10/07); other waste can be recycled or used as hardcore and levelling material. At Walmgate the non-structural evidence is sufficient to indicate smithing on the site, the 138kg of slag recovered perhaps suggesting that nearly 14 tonnes of artefacts were produced. This implies a substantial iron smithy on Walmgate (in Buildings U and W), the largest so far excavated in York. Structural evidence recovered included hearths, stake- and post-holes as well as rubble footings that may have supported raised bellows, possible waist-high furnaces, concentrated patches of rubble levelling perhaps to support a large wooden block on which an anvil would have been positioned, as well as a possible anvil slot and associated pits for quenching.

In conclusion, the metalworking evidence suggests a large-scale iron smithy with a secondary medium-sized copper alloy casting operation situated within a complex of timber-framed buildings and workshops, dating from the mid 14th to the late 16th century; evidence for earlier small-scale metalworking on the site or in its vicinity was also recovered. Cauldrons and other vessels of non-ferrous metal were cast alongside the iron being worked in the smithy. The evidence suggests that living and working quarters were in close proximity and that at times there was a shop on the Walmgate street frontage. The Walmgate evidence suggests that the Bedern/St Andrewgate area was not the only site of metalworking in York. The iron smithy at Walmgate may represent a suburban smithy, making goods for the local inhabitants of the area. Its products may have included horseshoes, nails, tools, structural fittings and knife blades. This important new evidence is a valuable addition to the growing corpus of urban metalworking in England in the medieval and early post-medieval periods.



Click to enlarge
Excavations in progress

Click to enlarge
Typical mould fragment

Click to enlarge
Furnace construction cut 2458

Click to enlarge
Mould fragment

Click to enlarge
Mould fragment

Click to enlarge
Crucible fragment

Click to enlarge
Medieval iron smithing
© Copyright York Archaeological Trust 2003