Group 23: Northumberland and the eastern Borders

This is the most straightforward of the Cosmographer’s groups in Lowland Scotland; it begins at Trimontium, known from Ptolemy (Geography II.3,6) and generally accepted as Newstead, overlooked by the triple peaks of the Eildon Hills. The third name is known epigraphically (RIB 1262 and 1270) and from the Antonine Itinerary Iter I to have been at High Rochester, so *Eburocastellum is probably to be identified with Cappuck, between the two places. Coccuueda should be a site close to Brinkburn Priory rather than a river-name, as *Coccuueda recurs for the River Coquet in the list of rivers at 108₃₈.

Alaunais almost certainly the fort at Low Learchild (Rivet & Smith 1979, 245), while the next name, which may be emended *Horrea Classis (Rivet & Smith 1979, 373), suggests a coastal site and should have been at the mouth of the River Tweed, where logic demands a Roman military presence, even if the Votadini were a tribe friendly to Rome. Steer (1958, 99) reports the discovery of an enclosure of Roman date at Tweedmouth, from which late third-century pottery was recovered. <Euidensca> is possibly an error for Habitanco (Rivet & Smith 1979, 371 contra Richmond & Crawford 1949, 34), a name known epigraphically from the fort at Risingham (RIB 1225).

<Rumabo>remains a problem: Rivet and Smith (1979, 240) and Dillemann (1979, 70) both suggest that this is an error for *Flum Abo, to be identified, according to the former, with the River Humber and Ptolemy’s Αβου ποταμου εκβολαι (II.3,4). Whilst the emendation is probably correct and the name (‘River’) certainly suggests a river of some size and importance, such as the Humber with which Rivet and Smith identify it, this is too far to the south to have been included here with any logic. The only river of sufficient size for which no Romano-British name is known in this area is the Tweed, a not impossible identification. The likely existence of a pre-English etymology for Tweed (Ekwall 1928, 423) is not a problem: the Humber, undoubtedly Ptolemy’s Αβου ποταμου εκβολαι, is also of probable Celtic origin (Ekwall 1928, 205). We should perhaps imagine a scenario similar to that sometimes proposed for the origins of the many Avons in England: the Roman official asks a local ‘what is that called?’, pointing to the river. The local, misunderstanding the question, replies ‘River’ or ‘Water’, which is duly logged by the official as if a name. The original name survives in local usage, while a new name has been coined for official purposes.