Attested: Durnonovaria on iter 15 of the Antonine Itinerary (also duplicated as part of an accidental insertion in iter 12). No epigraphic evidence has been found for this name, which does not appear in other texts where it might also be expected (Cosmography, Ptolemy).
Where: Probably Dorchester, around SY692905, an important Roman town in the valley of the river Frome. The Itinerary supplies a distance of 8 Roman miles from Vindocladia (Poole), which must be amended to 18 to fit Dorchester, the same change as needed 2 stages earlier in Iter 15. Then 36 miles to Moriduno (probably Gittisham) implies that the Roman road into Devon was straighter than Margary's route 4f suggests. Wareham is much less likely.
Name origin: It makes best sense to take out one letter N and amend the name to *Duronovaria, so that the first element becomes Duro- ‘crossing’. Then the second element -novaria may be compounded from *navis ‘river’ plus PIE *ar- ‘to fit together’, analogous with the two-rivers mouths of the Naver in Scotland and of the Naurum Taw plus Torridge, in Devon. An alternative analysis would separate out the -varia, to be analogous with the first element of Wareham, which is usually guessed to mean ‘weir’, though PIE *wer- had many possible meanings, among which ‘high spot’ is particularly plausible. There is also an argument for *var- or *ver- being an ancient river name, as in the rivers Var and Vire in France (Delamarre, 2003 p300). In that case, what did -nono add to the basic Duro- ‘door/through’ element? Should one take note of Old English durena ‘of the doors’ or durne ‘hidden’ or of ancient Onno ‘river’?
Notes: Rive & Smith followed a long tradition of taking out two letters NO, to amend the name to **Durnovaria, in order to liken the first element to Welsh dwrn ‘fist’ and to fit the remark by Bishop Asser, the biographer of Alfred, in 893, that the Welsh name for Dorset was Durngueir. Not convincing. The idea that iter 15 (analysed in RRRA Newsletter summer 2019) entered Dorchester along the known Roman road from Badbury Rings and Shapwick is problematic, for multiple reasons, but no Roman road from Poole has yet been identified. Roman Dorchester has a very long write-up in British History Online, but watch out that it begins with some glaring errors. Dorchester did develop into a walled Roman town, which took over from the huge hill fort of Maiden Castle 3 km south-west, but it has no known Roman fort. Its remarkable Roman aqueduct is a strong candidate to claim the difficult name apparently written as omire tedertis. The other Dorchester, in Oxfordshire, has no known Roman name, but it had a Roman river crossing of the Thames at SU587931, which was replaced by some fords that were strategically important in the Middle Ages.
You may copy this text freely, provided you acknowledge its source as www.romaneranames.uk, recognise that it is liable to human error, and try to offer suggestions for improvement.
Last edited 4 February 2021 To main Menu.