Cite record | Embed record | Print | | Report
BUCKLE
Unique ID: NMGW-1ED1A1
Object type certainty: Certain
Warning: getimagesize(./images/mlodwick/display/2009.218.1sm.jpg) [function.getimagesize]: failed to open stream: No such file or directory in /home/beowulf/public_html/app/views/scripts/partials/database/thumbs.phtml on line 47

View | Zoom | Download
Warning: getimagesize(./images/mlodwick/display/chepstow buckle a.jpg) [function.getimagesize]: failed to open stream: No such file or directory in /home/beowulf/public_html/app/views/scripts/partials/database/thumbs.phtml on line 47

View | Zoom | Download
Late Roman copper alloy buckle, probably of 4th or early 5th century date (Worrell 2010, 416, no. 1, fig. 3). The buckle is complete, comprising a frame, pin or tongue and plate (with an overall length of 97mm and a weight of 72.8g). The buckle frame is zoomorphic with two horse's heads facing each other and meeting at the centre (with a maximum thickness of 5.5mm). The ears of the horse protrude at the sides (producing a maximum width for the frame of 42.9mm). The snouts of the horses expand towards their ends and are separated by moulded grooves, defining a central rounded knop. The grooves accommodate the twin double-fork of the pin. The eyes of the horses are defined by cast spirals in relief. The spirals tail to form a rib along the necks of the horses, outside of which the frame is thinner (2.3mm compared to 4.2mm). At the base of the neck, at the attachment end of the buckle are three applied grooves, flanking a hinged loop at each side (with a depth of 8.7mm). The buckle frame has in-turned or involuted terminals. The rear of the frame is slightly convex with some finishing.
The pin or tongue has a wide cast loop (10.2mm external diameter) and gradually widens and thickens before forming rearward projections (with a width of 22.7mm), which complement the in-turned terminals of the frame. The tongue has a double-forked pin. The pin and buckle frame are secured to the buckle pate with a circular-sectioned axis bar (2.8mm diameter), which is secured by closed loops in opposing directions at the terminal.
The buckle plate (with an overall length of 66.2mm) has four hinge loops, which flank the two frame loops and the central tongue loop. The plate has openwork decoration within a rectangular frame (56.8mm long x 36.7mm wide and 3.6mm thick). The frame has an incised border with a rectangular panel on the frame side, containing diagonal grooves from alternating sides. At the centre of the frame is an openwork horse, trotting and facing towards the frame end. Grooves on the horse are used to define the snout, eyes, ears, mane and hoofs. Above the back of the horse is an pelta motif enhanced with grooved arcs. The pelta is thinner than the horse and it is unclear what, if anything it is intended to represent, possibly a tree or a much stylised rider. The rear end of the horse is poorly executed compared to the front. The reverse of the plate has four integral D-shaped lugs (8-9mm high and giving the plate a maximum depth of 12.8mm), positioned near each corner. Each lug has a central circular perforation (3.6mm - 3.9mm diameter) used to attach the buckle to the belt.
The buckle has good surface preservation; a recently applied coating by the finder has produced a dark-green patina.
Analysis to ascertain the elemental composition of the artefacts was carried out using a CamScanMaXim 2040 analytical scanning electron microscope (SEM), plus an Oxford Instruments Link Isis energy dispersive X-ray spectrometer (EDX).The objects were examined whole within the specimen chamber and were analysed for 100 live seconds using a working distance of 35mm and an accelerating voltage of 20kV. Where possible, areas of metal exposed from under the surface corrosion were analysed. Apart from very superficial surface swabbing of the artefacts there was usually no further sample preparation. The results are semi-quantitative and overall totals taken from minimally prepared surfaces are usually affected to some degree by the surface geometry of the artefacts, including factors such as curvature, pitting and indentations etc. Other factors that can also affect the results include surface oxidation, surface enrichment or depletion of certain elements, contamination from closely adhering dirt, and preferential corrosion.
Three separate readings were taken and the majority of the overall totals acquired were slightly low due to the factors described above. The results were normalised to 100% to obtain consistency between the readings and to make the analyses comparable with other results the average compositions were: 88-90% copper, 4% zinc and 5% tin. Minor amounts of lead, nickel and antimony also detected
The buckle type is generally characteristic of a type of late Roman buckle, sometimes referred as the 'Soldiers and Settlers' type after an article by Chadwick, Dunning and Hawkes (1961)[1]. The buckle generally fits into their type II A (ibid. p 50-51), although it is slightly different in having a pair of facing horses on the frame, rather than the more usual dolphins, with horses generally appear on buckles of type I B, with projecting horse's heads facing away from each other. The tongue rarely survives on this buckle type but when it does, it commonly has the rear-facing projections but rarely has a double pin. The axis bar with terminal loops is also difficult to parallel and may have been replaced or is a repair. The elaborate openwork plate is also fairly difficult to closely parallel. The central horse figure is reminiscent of 'Horse-and-rider' type plate brooches of 2nd century date. The best identified parallel for the buckle plate is an example from Argeliers in southern France, published by Aurrecoechea Fernandez (1999)[2] as his 'Pseudo-Hispanic' 'Dolphin' Belts with Figural Decoration: 'Santome' type. As noted by the type name, the Argelier buckle has dolphins on the frame but has a similar plate with openwork horse and is without the pelta shape and also the rear attachment lugs. Pseudo-Hispanic buckles are characterised by shanks on the rear to attach the belt and examples with perforated lugs are difficult to identify. Aurrecoechea Fernandez (ibid) dates this type to between AD350 and 380 on the continent.
[1] Chadwick, S., Hawkes & Dunning, G.C.; 1961; Soldiers and Settlers, fourth to fifth century: with a catalogue of animal-ornamented buckles and related belt-fittings; Medieval Archaeology; Journal of the Society of Medieval Archaeology, Vol. 5, p 1-70
[2] Aurrecoechea Fernandez J.; 1999, Late Roman Belts in Hispania; Journal of Roman Military Equipment Studies Vol. 10, p 55-62
This is a find of note and has been designated: Potential for inclusion in Britannia
Subsequent actions
Subsequent action after recording: Returned to finder
Chronology
Broad period: ROMAN
Date from: Circa AD 300
Date to: Circa AD 500
Dimensions and weight
Length: 97 mm
Width: 42.9 mm
Thickness: 5.5 mm
Weight: 72.8 g
Quantity: 1
Materials and construction
Primary material: Copper alloy [scope notes | view all attributed records]
Manufacture method: Cast [scope notes | view all attributed records]
Completeness: Complete [scope notes | view all attributed records]
Spatial data
Region: Wales
County: Monmouthshire
District: Monmouthshire
Parish: Chepstow
Restricted 4 Figure grid reference: ST5396
The map has been degraded and provides an approximate location with a degree of random obfuscation.
Grid reference source: From a paper map
Grid reference accurate to a 10 metre square.
Method of discovery: Metal detector
[scope notes]
Personal details
Found by: This information is restricted for your login.
Recorded by: Mr Mark Lodwick
- [
view all attributed records]
Identified by: Mr Mark Lodwick - [view all attributed records]
Secondary identifier: Ms Sally Worrell - [view all attributed records]
Other reference numbers
Other reference: NMWPA 2009.218.1
References cited
No references cited so far.
QR barcode
If you have a mobile phone equipped with QR recognition software, you can go directly to the webpage that this record resides at. Every record has an individual one of these.
Spotted a mistake? Tell us. | Be the first to comment
Comment on this artefact's record
Data entered via this form is checked against the akismet service to recognise spam.
Audit data
Created:
Tuesday 30th March 2010
Updated: Friday 24th February 2012

