Bruce O The one I mentioned was part of the Scottish Templar argument - I find reading that sort of thing a form of relaxation.

Pete M, some more names More Kentish Celtic. Coed also appears in Chatham and Chattenden, on the Medway (second element cognate with Welsh gwy, winding), also a Celtic name, as is the Cray, for some time the boundary between Britons and Jutes. Lympne and Lyminge, (though the ge suffix is English) carry the name of the Limen, the old name of the Rother, meaning elm. Thanet may have derived from a word cognate with Welsh tan, fire. Richborough was formed from a contraction of Rutupiae through Repta to Ratteburg and thence Richborough (possibly changing to avoid a rodent connection). Blean is apparently British. There are some minor placenames such as Wycombe, which may incorporate cwm.

There are also Roman survivals in names with wick (from vicus), and various combes, and comps (from camp). Faversham may be from faber, and there is an argument for a Roman workshop surviving there into English times. Eccles is argued about. It either was, or was not, an early Christian ecclesia.

One book points out that there is a swathe of Romano-British names round the coast fom Pevensey to London, and that there has to be a common cause for these survivals, such as their being known to settlers before their arrival.

There's archaeological evidence for the continuity of pre-Roman estate boundaries down to parishes in the present, for example at Lullingstone, where the villa was succeeded by an English settlement, which also argues for some sort of continuity of society of some sort.