Group 6: southern Hampshire/northern Wiltshire

Group 6a:southern Hampshire



10617 = 10620




Venta uelgarom









Nauimago regentium


10620 = 10617

This tight group takes us on a roughly circular tour of south-east Hampshire and western Sussex. Having jumped to the no doubt prominently marked Civitas capital at Noviomagus *Reginorum, Chichester, the Cosmographer travels north to Onna, probably the ‘small town’ at Neatham or, less likely, the posting-station at Iping (Rivet & Smith 1979, 432), and then west to Winchester, Venta Belgarum. The next two names are rather more difficult: Ardaoneon has been connected with the Portum Adurni of the Notitia Dignitatum (Occ. xxviii.21) and identified with the fort at Portchester (Rivet & Smith 1979, 442), while Armis has been taken as the name of the River Erme, far to the west in Devon (Rivet & Smith 1979, 259).

This latter suggestion is most improbable as Ekwall (, 149) gives a convincing Old English derivation for Erme, although the identification of *Ardaunium (Richmond & Crawford 1949, 23) with Portchester can stand, and its connection with the name in the Notitia is probably correct. It is interesting to note that *Ardaunium, alone of all the Saxon Shore forts, has the epithet Portus in the Notitia Dignitatum, and that this seems to have survived in the modern names Portchester and Portsmouth, whereas the Celtic form (meaning something like ‘The Height’, probably referring to Ports Down) has disappeared. The derivation of Armis from a Celtic word meaning ‘Arm (of the sea)’ (Rivet & Smith 1979, 259) ideally suits the promontory site at Bitterne, which cannot on this hypothesis be the Clausento of the Antonine Itinerary Iter VII; the latter is perhaps therefore a hitherto unrecognised site at Wickham, as the mileages from Chichester and Winchester can be taken to suggest (Rivet & Smith 1979, 166).

Finally the Cosmographer returns to Nouiomagus *Reginorum and names it again in error, this time with its tribal epithet. The correct form of the tribal name has been much discussed since Haverfield (1912, 348) suggested the emendation *Regnenses, based on the <regentium> of the Cosmography. Although many of the more traditional historians of Roman Britain have been loath to give up this emended form because of its associations with the kingdom (regnum) of Cogidubnus (e.g. Frere 1987, 192), Celticists have been emphatic on the point that we are dealing with an original *Regini and its variant Regni, meaning ‘the proud people’ (K H Jackson in Rivet 1970, 78; D Ellis Evans in Rivet & Smith 1979, 446). True enough, it is a remarkable coincidence that a regnum should have been created which included a tribe called *Regini (Frere 1980, 422), but this may have been added stimulus to assimilate a British Celtic word to a Latin one.

Group 6b: north Wiltshire


East Anton





This pair of names does not properly fit with the previous group or the following (which cluster around the Severn estuary) and seems to form a link between them. Returning to Venta, Winchester, which is not named again, we move north-westwards along the road to Cunetione, Mildenhall (Antonine Itinerary Iter XIV), which shows assibilation of -ti- to -tz- in the Cosmography, via Leucomago, in all probability the settlement at East Anton, near Andover (Rivet & Smith 1979, 389). Richmond and Crawford (1949, 37) successfully predicted the existence of this settlement, demonstrating that the Cosmographer can be relied upon to follow a fairly orderly listing of names on occasion.