Given the vast numbers and varieties of brooches recovered from Roman Britain, evidence for their manufacture is surprisingly sparse.Footnote 42 The bulk of this evidence comes from the south of the province, although distributional analysis makes it clear that the frontier zone saw a thriving tradition of brooch production.Footnote 43 This note draws attention to some Scottish evidence — one a recent find, the other long published but overlooked in recent surveys — which raises interesting questions about the organisation of brooch production in the military zone.
STEVENSTON SANDS, AYRSHIRE (FIGS 5–6)
Among a collection of antiquities donated to the National Museum of Antiquities of Scotland in 1931 on the death of the antiquarian John Smith were two Roman brooches from Stevenston Sands, Ayrshire.Footnote 44 They have been published previously, but the wider significance of one has been overlooked.Footnote 45 Both are enamelled plate brooches. One is a lozenge-shaped brooch with a central cup, now empty, surrounded by four enamelled fields alternating between pale blue tips and opaque red sides, with eight small reserved bosses; it has a single lug for a sprung pin on the reverse ( fig. 6a).Footnote 46 Bayley and ButcherFootnote 47 date this type to the second to early third century a.d., and suggest a Continental origin. Of interest here, however, is the second brooch ( fig. 6b). This is a lunate or peltate plate brooch with knobbed terminals.Footnote 48 The front has two empty curved fields prepared for champlevé enamel. On the reverse, fittings for the pin are unfinished; the catch-plate has not been turned to hold the pin tip, and the hole for the spring has not been drilled in the single fastening plate, though its position is marked with a dot. This is a casting which was cleaned up but not completed. MackrethFootnote 49 quotes second-century parallels, though few are closely dated; the distribution is widely dispersed.
Findspots of the two unfinished brooches.
Brooches from Stevenston Sands. (a) Lozenge plate brooch; (b) Unfinished peltate brooch. (Drawn by Alan Braby)
The finder, John Smith, was an amateur archaeologist, geologist and botanist who carried out extensive research into the archaeology of Ayrshire, publishing a major survey of the area's prehistoric archaeology and excavating a number of sites.Footnote 50 He maintained an extensive private collection, derived from his own excavations and field collecting, which was donated to the National Museum on his death.Footnote 51 Stevenston Sands (also known as Ardeer Sands) are an extensive area of sand dunes on the Ayrshire coast, well-known for their archaeology.Footnote 52 Unfortunately, as with so many dune collections, our knowledge comes entirely from casual discoveries; in the absence of excavation little is known of the nature of the archaeology in these dunes, and it is not possible to offer any precise provenance for the brooches, or even say if they were found together. However, the large quantities of later prehistoric finds from the area point to considerable activity in the Iron Age, continuing into the Roman period.Footnote 53 While the precise context remains vague, this discovery provides the most northerly evidence so far for brooch manufacture in Roman Britain.
FALLA KNOWE, ROXBURGHSHIRE, SCOTTISH BORDERS (FIGS 7–8)
The other discovery is a more recent find, one of a number of objects recovered by Mr W. Butler while metal-detecting around an Iron Age homestead at Falla Knowe in 1997 and 2001 ( fig. 7). The site appears typical of late Iron Age/Roman Iron Age settlements in the area, comprising an irregular sub-rectangular enclosure made up of four subordinate enclosures containing a huddle of roundhouses and adjoining yards.Footnote 54 Near by are a linear earthwork and traces of possible field-systems, while Dere Street runs some 350 m to the east. The nearest known military sites are the temporary camps at Pennymuir, about a kilometre to the south-east; the nearest known permanent installations are the fortlet of Cappuck, some 8 km to the north-west, and the signal station on Brownhart Law, around 7 km south-east.Footnote 55
Location of Falla Knowe. On 7B, 1 = the findspot of the bridle bit link; 2 = the brooch and mount. (Drawn by Alan Braby)
Three copper-alloy finds were recovered: a brooch, a strap mount and a bridle bit fragment. The key object is an unfinished plain trumpet brooch with full petal moulding (reduced on the underside) flanked by triple ridges ( fig. 8a).Footnote 56 It has a tightly curved head, with no integral headloop nor any surviving lug for a wire one; the foot and catchplate are missing from recent damage, but it seems the foot was slightly twisted. It is a poor-quality unfinished casting; the casting seam is clearly visible and the two sides of the mould were slightly misaligned, creating a step along the seam. The plate under the head to hold the spring has not been perforated, indicating the casting was discarded before finishing.Footnote 57
The other two objects are typical of the local Roman Iron Age; both are styles familiar from local metalwork, although they are also found in Roman forts. One is a strap mount with boss-and-petal decoration ( fig. 8b).Footnote 58 The other is the central bar of a derivative three-link bridle bit ( fig. 8c), with a central swelling and two oval attachment rings, one worn through, the other with what is probably part of the cheek-ring corroded onto it.Footnote 59
Falla Knowe finds. (a) Unfinished trumpet brooch; (b) Harness mount; (c) Bridle bit link. Scale: (a and b) 1:1; (c) 3:4. (a & c: drawn by Marion O'Neil; b: by Alan Braby)
The bit fragment came from the north corner of the site, in material collapsed from the outer face of the enclosure wall, while both brooch and mount came from within the enclosure, just west of the largest scooped building, beside a low rectangular platform which is not shown on the published plan. Magnetometer survey of the siteFootnote 60 revealed an area of enhanced magnetism around the brooch findspot, perhaps a metal-working area.
DISCUSSION
These two brooches are valuable additions to the evidence for brooch manufacture in Roman Britain. The evidence of trumpet brooch manufacture in this area is not unexpected: Mackreth has argued for the type's origins in the Midlands, but it was very popular across a spectrum of sites in the military zone, with evidence of production in the hinterland of Hadrian's Wall.Footnote 61 The lunate brooch is more unexpected, as it is not otherwise known in Scotland, but examples are reported from Hadrian's Wall.Footnote 62 It is noteworthy that neither of these brooches comes from a military site. The nature of the Stevenston Sands findspot is unclear, but undoubtedly non-military; given other finds from the area, it could be from an indigenous settlement or could represent an itinerant metal-worker (as discussed further below). By contrast, the Falla Knowe site is typically local in character; such enclosed scooped homesteads are characteristic of the later Iron Age/Roman Iron Age settlement pattern of the area, and the pieces of indigenous-style horse harness confirm this.
The location of the Falla Knowe site is intriguing, and raises further possibilities. Its proximity to Dere Street might suggest that production was targeted towards traffic moving along the road. The intended audience for the Stevenston Sands producer is less clear; our knowledge of the military presence in Ayrshire is very partial indeed, although a fort site is likely at nearby Irvine,Footnote 63 but such brooches would have been equally desirable in a local context. This brooch owes nothing to local traditions, which might suggest it is better seen as the product of an itinerant craftworker from further south.
These finds emphasise the complexities of studying interaction in the frontier zone. When found on indigenous-tradition sites, Romano-British brooches are generally treated as imported items,Footnote 64 but this evidence shows some were being made locally on such sites. There was a strong selectivity in the brooches on indigenous sites, with a marked preference for types such as trumpet, headstud and dragonesque which resonated with existing local tastes, or for more ornate, often enamelled types, as with the Stevenston Sands example.Footnote 65 As Collingwood noted long ago,Footnote 66 such Romano-British types grew up in the interaction between local craftworkers and incoming military culture in the frontier zone, and represent a fusion of indigenous and introduced styles. It is thus entirely feasible to find such Romano-British brooches being made on indigenous sites in the occasionally-occupied area between the Walls, although the habit of brooch-wearing itself had been exceedingly rare in the Scottish Iron Age, and seems to have expanded greatly under Roman influence.Footnote 67
ACKNOWLEDGEMENTS
I am grateful to Billy Butler and Walter Elliot for their assistance on a field visit to Falla Knowe, to Diane Oliver for access to the site, Dave Cowley for information on recent survey data, the Trimontium Trust for sponsoring a geophysical survey and CFA Archaeology for carrying it out. XRF analyses were conducted by Katherine Eremin; illustrations are by Marion O'Neil and Alan Braby. An earlier draft benefited from the comments of an anonymous referee.
