[NB. The reader will probably find it easier to understand this page if he or she has already acquainted himself or herself with the contents of Chapters 1 and 2. Chapter 1 concerns the basic building-blocks used in Romano-British place-names and Chapter 2 explains how those building blocks were assembled by the Celts to form compound names. The numbers accompanying some names are those provided by Richmond and Crawford in Richmond and Crawford 1949.]
Cironium to Sandonio
From Cirencester to the river Trent, via Dover and Caernarfon
1 The above is not intended to be a funny title, but a description of what Ravenna now does, for starting with Cironium Dobunorum (66) at Cirencester the compiler of Ravenna takes us away over to Dover, then sweeps back again to Caernarfon and then goes back east again to Sandon, on the river Trent, picking up Romano-British place-names on the way. The name Cironium includes the element cir meaning ‘steep hill’, which is not really suitable for Cirencester itself, but is suitable for Bagendon, a little to the north. It is known that there was a Dobunni settlement at Bagendon and it is thought that the Romans re-settled the Dobunni at Cirencester, so it may simply be the case that the Romans adopted the name Cironium of the Dobunni settlement at Bagendon and applied it to their fort and new settlement at Cirencester. The first stop after Cironium is Caleba Arbatium (67) at Silchester, and this presents exactly the same problem as Cironium. The cal of Caleba means ‘steep hill’, which is not really appropriate for the actual site of Romano-British Silchester. It would seem, however, that the form Calleva is correct (see the entry for Caleba in the Alphabetical List), so the name will simply have been transferred from the Iron-Age fort at Pond Farm, about one kilometre northwest of the Romano-British town. That fort stands on a fairly steep slope and so will have had a name of the form Calveva or Calveba originally (the element calv meaning ‘steep hill slope’), this then changing to Calleva (cf. Banva → Banna).
2 The next Ravenna name is Anderelionuba (68), apparently at Pevensey. Note that Anderelionuba is actually a river-name transferred by the Romans to the fort at Pevensey. The name is of the kind which has a river element, in this case uba, including the river-letter b, added as a suffix to a place-name, in this case Anderelion. We can thus deduce that there was a place called Anderelion on the banks of that river. We can’t say where it was, but we do know that the earliest settlement was on the top of raised ground, this being the meaning of the and element. The next name in Ravenna is Mutuantonis (69). This is a topographical compound with two hill-letters, so the name is that of a Celtic settlement which existed prior to the Roman invasion, and we know from the order of names in Ravenna that that settlement was between Anderelionuba at Pevensey and Lemanis at Lympne. In the absence of any clear alternative Mutuantonis is here identified as the Celtic settlement on the East Hill in Hastings. The settlement was on the summit of the hill, so the Celtic name was presumably Mutuandonis, comprising the inversion-type element Mut, meaning 'hill high', qualified by the old-style element and, meaning 'hill summit'. There may have been a c meaning 'steep' between the u and the a of the Celtic name. The Ravenna name may of course refer to a Roman post which took its name from the Celtic settlement - that post may have been built inside the old settlement or somewhere else in the vicinity. Then come Lemanis (70) at Lympne, Dubris (71) at Dover, Duroaverno Cantiacorum (72) at Canterbury and Rutupis (73) at Richborough. Dubris is a modified form of the river-name Durbis seen in Ravenna and Rutupis a modified form of the river-name Turupis. The bis of Durbis, changed to pis in Turupis, will just be a name-ending, and Dur/Tur comprises the river-letters t and r corresponding to the hill-letters l and m, as seen in Lemanis. The river-letters t and r are also seen in the ther part of the modern river-name Rother - the river Rother apparently reached the sea at Lemanis/Lympne before the Royal Military canal was built. The river-names Durbis and Turupis were transferred by the Romans to forts/harbours built at Dover and Richborough respectively, though at some point the names were changed to Dubris and Rutupis, probably just to make it easier to distinguish one from the other. But the river-names were not changed - the Dur of Durbis is now the river-name Dour and the Tur of Turupis is now the tour part of the river-name Stour. The next name is Durobrabis (74), at Rochester. The abis of Durobrabis is just an ending and br is an old-style element meaning 'high hill'. The Celtic settlement called Brabis will have been up on the high ground on the eastern side of the Medway, somewhere in the vicinity of the later Romano-British town of Durobrabis.
3 But after Durobrabis Ravenna causes a slight difficulty by listing Landini (75) and Tamese (76). Landini or Landini Tamese (a name like Stoke-on-Trent or Newcastle-upon-Tyne) might be an early name for London, though London is listed later as Londinium Augusti (97). Another possibility is that Landini was at London and Tamese was a quite separate place, most probably at Streatley. The third possibility is that Landini Tamese was at Streatley. The point is, as explained in Chapter 12, that Ravenna appears to have started a journey from the Thames valley to South Wales at Leckhampstead (Leucomago). It seems more likely, however, that it would actually have started the journey at the important crossing of the Thames at Streatley, but the compiler of Ravenna didn’t give the name of the fort at Streatley at that point because he planned to list it later. Landini Tamese would then be a fort up on the high ground on the west bank of the Thames at Streatley. All this would mean is that the compiler of Ravenna had used Silchester as a node, with one route out of Silchester going over to Rochester via Dover and the second heading north from Silchester to Streatley. Streatley lies on the road from Silchester to Brinavis (77) at Bicester, the next port of call. Brinavis appears to be a straightforward old-style compound in the hill-letters r and n. The element br means ‘high hill’, a clear reference to the nearby Graven Hill.
4 Ravenna now goes in a straight line to Vtriconion Cornoviorum (79) at Wroxeter. And at the point where that straight line crosses the river Avon, at Stratford-on-Avon, lay Alavna (78). Alavna is of course a river-name. The Romans simply adopted it and applied it to a fort which they built on the banks of the river at Stratford. The spelling of Vtriconion is presumably just an error – it should be Viriconion. But then Viriconion Cornoviorum presents the same problem as Cironium Dobunorum and Caleba Arbatium. The viric element appears to be an inversion-type element meaning ‘slope of hill steep’, but if the original spelling was biric then this is a transitional element meaning ‘high hill steep’. It seems most likely that Biriconion was the hill-fort of the Cornovii on the Wrekin and that the Romans simply adopted the name of that settlement - later changed to Viriconion - and applied it to their fort at Wroxeter. The Cornovii were presumably moved from their old settlement to a new settlement next to the fort.
5 After Wroxeter Ravenna lists Lavobrinta (80) and Mediomano (81). Lavobrinta, with the spelling changed to Lacobrinda, is a perfect topographical description of the location of the fort at Caer Gai. The lac element is an inversion-type element meaning ‘hill steep’, and brinda is an old-style compound in the two elements br, meaning ‘high hill’, and ind meaning ‘hill summit’. The latinisation of the c in Lacobrinda to the v of Lavobrinta is seen in several other place-names, for example Leviodanum (220), Levioxava (222) and Ravatonium (209). Mediomano is, however, a puzzling name, in that one would not expect the name to include the hill-letter m twice. The fort in question is that at Tomen-y-Mur and it is possible that Tomen is derived from the dioman part of the old name. Perhaps the first m was originally some other hill-letter, n being perhaps the most likely (though s is also possible), but whatever the hill-letter was the meaning of the first element of the name is ‘hill summit’. The name is thus entirely appropriate for the fort at Tomen-y-Mur. After Tomen-y-Mur Ravenna takes us to Seguntio (82) at Caernarfon. As it stands, the name comprises two inversion-type elements, seg meaning ‘hill steep’ and unt meaning ‘hill high’. However, since the fort was actually built on top of the hill it is probable that the original Celtic form was Segundium (but see the entry for Seguntio in the Alphabetical List).
6 After Seguntio at Caernarfon comes Canubio (83) at Caerhun. The name Canubio is simply a shortened version of the river-name Descecanglubena, an alternative name for Ptolemy's Tuerobis river, the modern river Conwy (this is discussed in more detail in the notes for Ratostabius to Toesobis, which may be accessed from the table in paragraph 9 of Chapter 19). The Romans transferred the modified river-name to their fort at Caerhun. Ravenna now lists Mediolano (84) at Whitchurch. Mediolano is often taken by scholars to mean ‘in the middle of a plain’, or ‘central plain’, both of which seem rather far-fetched. The fort at Whitchurch was up on top of a hill, which is actually the meaning of the med element. The l is the hill-letter l and the ano just a name-ending. After Mediolano comes Sandonio (85) at Sandon, on the river Trent at a point to the northeast of Stafford. The sand element is like the lond of Londinium, but using the hill-letter s rather than l – the element appears to be an old-style compound in the hill-letters s and n, where the nd element indicates that the Celtic settlement was on the top of raised ground. One would therefore expect to find traces of that settlement on the top of raised ground in the vicinity of the modern village of Sandon. The Roman fort may have been on the same site or possibly somewhere else in the vicinity, perhaps close to the Trent to control a crossing of the river. The place-name will then have been transferred by the Romans from the Celtic settlement to their own fort. Note that Trisantona, now the Trent, may be a river-name of the kind having a river-element, here a compound of the river-letters t and r, used as a prefix to a land-name, here Santona, presumably a slightly modified form of Ravenna’s Sandonio at Sandon on the Trent. It is of course alternatively possible that Trisantona is a straightforward river-name in the river-letters t (corresponding to the hill-letter l1), r (corresponding to the hill-letter m), s (corresponding to the hill-letter r) and t (corresponding to the hill-letter l2). The Roman place-name Sandonio may then be derived from the river-name.
7 Ptolemy mentions some of the Ravenna names discussed above. He mentions Corinium at Cirencester, Caleva at Silchester, Daruernum at Canterbury and Rutupie at Richborough. He also mentions Londinium at London and assigns it to the Canti tribe, though there are scholars who see this as a mistake on Ptolemy’s part. And he mentions Viroconium at Wroxeter and Mediolanum at Whitchurch.
8 The locations of the forts or settlements discussed in this chapter are shown on the maps below. The forms of the names and the numbering of the names are those given by Richmond and Crawford. The identifications are as follows:
66 Cironium Dobunorum Cirencester
67 Caleba Arbatium Silchester
68 Anderelionuba Pevensey
70 Lemanis Lympne
71 Dubris Dover
72 Duroaverno Cantiacorum Canterbury
73 Rutupis Richborough
74 Durobrabis Rochester
75,76 Landini Tamese probably Streatley, but possibly London
77 Brinavis Bicester
78 Alavna Stratford-on-Avon
79 Vtriconion Cornoviorum Wroxeter
80 Lavobrinta Caer Gai
81 Mediomano Tomen-y-Mur
82 Seguntio Caernarfon
83 Canubio Caerhun
84 Mediolano Whitchurch
85 Sandonio Sandon on the river Trent
Mutuantonis at Hastings is not shown on the maps. It was between Anderelionuba and Lemanis.
[This page was last modified on 08 March 2021]