The Didcot (formerly Oxfordshire) Mirror

Didcot (formerly Oxfordshire) Mirror
photo stated as "free use" since 2009
(image rotated)

Celtic decorated mirrors are the quintessential British La Tène object. Only rarely found with other objects, they are notoriously difficult to date with precision but their date range is from before the middle of the first century BC to about the middle of the first century AD. This one (No. 23) was found by a metal detectorist in 2007 and reported to the Portable Antiquities Scheme, although it does not seem to be on their site after a search for "mirror" or "Didcot".  The handle was found in three pieces and the mirror was restored at the St. Albans Museum.

Decorated mirrors fall into two broad chronological divisions: the earliest (such as this one) have a round plate with no binding on the edge of the plate. The later mirrors have a kidney-shaped plate with a binding strip around the edge of the plate.

They also have two dominant plate decoration types: first is the "diffused" type, the other is "tri-lobed", often resembling an abstracted woman's head.

In order to better define these mirrors, two parts of the mirror must be taken into consideration: the design of the handle, and the applied decoration to the plate.

Shillington, Bedfordshire, Mirror
photo stated as "free use" since 2009

The closest parallel, taking into consideration both features, is the Shillington, Bedfordshire mirror (No 26) which is another metal-detector find that was found in November, 2000, in association with a silver knotenfibel brooch and pottery shards, and was thus dated 70 - 20 BC.

The main difference between the two mirrors is that the on the Shillington mirror, within the delineated decoration, are circled tricorns, whereas this element is replaced, on the Didcot mirror with roughly circular shapes in association with pairs of opposed diagonally set leaves.

The earlier plain mirrors from the Arras and Garton Slack burials in Yorkshire have, respectively, a straight handle with a ring at the bottom, and a double-baluster handle with median band and a ring at the top and bottom. In the Didcot and Shillington mirrors, and in the plain mirrors from Portland, Dorset (Jope, 2000, plate 256,c) and Bulbury, Dorset (ibid, plate 256,d) the top of the handle is bifurcated showing, I think, influence from earlier anthropoid-hilted swords with the top ring indicating the head and the bifurcated top section of the handle, the arms. On the Bulbury handle, the bifurcation does not end in "arms", but appears to continue to form the plate edge, perhaps a later development.

The development of the

anthropomorphic sword type

© Trustees of the British Museum

The development of the anthropomorphic swords is seen to the right. Note the occasional use of a median band. Another direction that the handles take can retain the bifurcated top but omits the top ring. The most famous example of this is the Mayer mirror (No. 20) where the arms taper wider at their terminals and the metal extends to form the start of the binding around the kidney-shaped plate. This seems to me to be possibly an early version of the late type of mirror, but chronological matters must also be tempered by the existence of different workshops and skill levels for roughly contemporary mirrors.

The other direction that the handle design takes is a series of rings, or a pair of opposed loops joined by a band, the latter seen, most famously, on the Birdlip and Desborough mirrors.

The use of opposed leaves in the mirror plate design is quite rare but can be seen (in association with a couple of circular forms) on the Bromham, Bedfordshire, Mirror (No. 6, and illustration below). It has the earlier round plate and no binding.

Bromham Mirror
 designated as "free use" since 2009

Most interesting of all of these examples is the smaller mirror from Dordrecht, Netherlands (No. 13) and illustrated below. It shows influences from both broad types of handles, and has a roundish plate leaning toward the kidney shape with (apparently) traces of binding. Here, the diagonally opposed leaves have a circle between, a design different, but reminiscent of one of my strap junctions that I date to about 20-40 AD. The plate design is a rather "diffuse" version of the "tri-lobed" type

Smaller Dordrecht Mirror 
photo stated as "free use" since 2009

Taking all of this into consideration, I place the date of the Didcot mirror more in keeping with the Shillington Mirror, but at the end of the cited period of 70 - 20 BC. or possibly even slightly later. Such an estimation must remain provisional for the time being because of the different workshop/varying skills factor.