In 2005 and 2006 Pre-Construct Archaeology carried out an excavation. A pit and possible building platform were dated to the early medieval period. The platform was constructed from reused Roman material. A gilded bow brooch cast in silver with moulded relief decoration was found in association with this feature. (1)

The brooch measures 84.7mm in length and weighs 31.97g. It has a semi-circular head-plate with three projecting knobs, each knob comprising three circular lobes in a triangular arrangement. The central panel of the head-plate carries curvilinear relief decoration. The curved bow is straight-sided with a flat cross-section. A central panel running down the length of the bow is inlaid with niello in a repeating motif of circles and lines known as a 'paragraphenmuster' ('paragraph') design. Either side of this is a panel of relief decoration comprising seven rectangular zones containing opposed semi-circles. The lozengiform footplate carries a central lozengiform panel of relief decoration in the form of a rosette, surrounded by an outer panel of semi-circular relief segments with raised dots. Two opposed openwork birds project either side of the upper footplate. Below these the edge of the lower footplate on either side is formed from a series of five semi-circular segments, creating a scalloped effect. At the terminal of the footplate is a circular projection. On the reverse the pin and spring mechanism, which is of iron, is complete (although the pin is now broken). The pin-bar passes through three evenly spaced lugs on the back of the headplate. The catchplate, which would have projected from the upper part of the back of the footplate, is missing, although close inspection reveals where it was attached. The front of the brooch is gilded, and the gilding shows very little sign of wear, perhaps implying that the brooch was only in circulation for a short time.
Relief
Date: from Circa 430 AD (Certain) to Circa 500 AD (Certain)
Period: Early Medieval
Broad Period:
Method of manufacture: Cast
Surface treatment: Gilded
Length: 84.71 mm
Width: 41.45 mm
Thickness: 4.36 mm
Weight: 31.97 grams
Completeness: Incomplete
OS GridRef: TQ79306850
Easting: 579400
Northing: 168600
Date found: 16 February 2006
Methods of discovery: Metal detector during controlled archaeological investigation
Circumstances of Discovery: Pre-Construct Arcaheology excavation. Found in fill of rectangular pit.
Recorded by: Andrew Richardson
Primary Identifier: Marit Gaimster
Subsequent action: Returned to finder
County: KENT
District: MEDWAY
Parish: GILLINGHAM
Address: Grange Farm
Known as: Gillingham
General landuse: Built over
Specific landuse: Other
PAS notes: The closest parallel to the Gillingham brooch is a very similar gilded-silver example with semi-circular head-plate that has been in Canterbury Museum since the late 1800s (Bakka 1958, 9, fig.2; Richardson 2005, II, 16, fig.2). This has long been regarded as a Scandinavian import decorated in the Nydam Style, an art-style that represented a transition between late Roman motifs and the Scandinavian style known as Style I. On stylistic grounds the Canterbury brooch has generally been dated to around the third quarter of the 5th century. However, as something of a 'one-off', certainly within England, it has not received a great deal of study. In addition, its findspot is unknown, although its presence in Canterbury Museum suggests that an east Kentish provenance is likely.

The Canterbury brooch measures 84.4mm long, and weighs 34g. Its overall form is identical to the Gillingham brooch, but it has differences of detail in the decoration on the head, bow and foot. Notably, the lower footplate is flanked by crouching animals (probably dogs), their heads facing downwards and their open mouths touching the terminal of the footplate, a drop-shaped projection depicting a human face-mask. This arrangement is directly derived from late Roman chip-carved metalwork. On the reverse, the pin/spring mechanism is of corroded iron and is also attached via three lugs on the back of the head-plate. Crucially, there are traces of textile preserved by contact with the iron corrosion, which strongly implies that this brooch is from a burial, since it must have been attached to a garment when it was buried. In addition, the gilding on the front of the brooch is very worn, perhaps suggesting a longer period of use than the Gillingham example.

The Canterbury brooch is regarded as being an import from southern Scandinavia. Superficially similar brooches are known from both southern Norway and Gotland in Sweden. However the rosette design on the footplate only appears on Anglo-Saxon, and some Continental, brooches, but never in this context on Scandinavian pieces. There are two other 5th-century brooches from Norfolk with a similar footplate (see The Searcher no.175, 22) and a Jutish-Kentish square-headed brooch from the Anglo-Saxon cemetery at Gilton, Ash, also carries a rosette on the footplate. The 'paragraph' decoration of the bow on both the Gillingham brooch and Canterbury example is virtually unparalleled in Scandinavia (and indeed more commonly occurs on Continental brooches of the 6th century). In itself, the 'paragraph' design is another late Roman design; it appears to have been initially used on weapon accessories such as mounts and buckles. there are also 5th-century Continental and Anglo-Saxon male burials with this feature. Given that both brooches discussed here were found in east Kent, and that in the later 5th- and 6th- centuries a distinctively Kentish workshop produced, under strong Scandinavian and Frankish influence, a series of square-headed brooches, the possibility that the Gillingham and Canterbury relief brooches are early east Kentish products must be considered. Whether they were in fact Scandinavian imports, or products of east Kent, they strengthen the picture of high-status Scandinavian artistic and cultural influence on east Kent during the second half of the 5th century.

Brooches such as these would have formed part of the costume of high-status women. The larger forms of square-headed brooches were usually worn as a single brooch, and it seems likely that the Gillingham and Canterbury brooches were also worn in this fashion. In contrast, in the 6th century, the Kentish series of square-headed brooches tended to be somewhat smaller than their Scandinavian or Anglo-Saxon counterparts, and were usually produced and worn as near-identical pairs or trios as part of a costume that usually also included a pair of circular (disc or button) brooches. Given their east Kentish context, their size and their probable status as single finds, a 5th, rather than 6th century, date of manufacture (and in the case of the Gillingham example, of deposition) seems probable. If they can be assigned to the Scandinavian Nydam style, this would place their manufacture in the period circa AD 450-480. However Karen Høilund Nielsen of Denmark suggests that whilst they are related to Nydam Style, they are best regarded as early Style I (pers. comm.), which would place their manufacture between circa AD 475-490.
Treasure case 2006 T78. Report by Marit Gaimster and Andrew Richardson.

<1> Pre-Construct Archaeology, 2008, Grange Farm, Gillingham, Kent: Assessment Of An Archaeological Excavation (Unpublished document). SKE17791.