[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 below are those provided by Richmond and Crawford in Richmond and Crawford 1949.]



Chapter 15

Navione to Alavna (187)


Roman place-names in the North of England


Part 2: Brocara to Alavna



[The text below is a continuation of that in Part 1: Navione to Fanococidi. Part 1 may be accessed by clicking here]


9     Then, after Maia Fanococidi, Ravenna lists a large group of names starting with Brocara (156). Since Ravenna has already listed names in northern England, and since four of the names in the Ravenna group are very similar to names assigned by Ptolemy to the Selgovae, it has long been assumed that the names in the Ravenna group all refer to places in southern Scotland. All manner of strange meanings have been proposed for individual ones of those names and any number of possible locations have been put forward. But it is notable that the survival rate of those names is remarkably low. The answer to this little mystery is that the names do not in fact refer to places in southern Scotland, but to places in northern England, and several of the names have survived, at least in part, in modern names. There is an error in Ptolemy in the names assigned to the Selgovae, though the blame cannot be laid at the door of Ptolemy himself, this point being discussed in Chapter 18. It would thus appear that Ravenna used at least three different maps for the north of England, the first being a Flavian map and showing the names from Navione up to Corielopocarium and the second a Trajanic map showing the names from Serduno to Maia Fanococidi. The third map is of uncertain date, though it may be even earlier than the first map. The first name taken from the third map is Brocara (156) at Brougham Castle, the name having the form Brocavo in the much later Antonine Itinerary (i.e. later than the map used by the compiler of Ravenna). Brocara is used as a node, the first two following names being Croucingo (157) and Stodoion (158). The oion ending of Stodoion is rather odd – perhaps the name was originally Stodonion. This may have been the fort at Nether Denton, it being possible that the den of Denton is derived from the don of Stodonion, just as the dun of Dunning is apparently derived from the don of Lodone and the dun of Dundon from the din of Lindinis. Note that the name Stodonion refers to a location on top of a high hill (st = ‘hill high’; don = ‘summit of hill’), a definition which seems entirely appropriate for Nether Denton.  Between Brougham Castle and Nether Denton are a river and village called Croglin and it is just possible that this name is derived from Ravenna’s Croucingo. But Croucingo is also an odd name – the adjectival c/g appears three times, but there are only two hill-letters, namely r and n. It is probable that the name is really an old-style name with three old-style elements, so it is possible that the Celtic name was somewhat of the form Corucingolonion, this form having been reduced to Croucingo in Ravenna and having survived as the modern Croglin.

10     But then Ravenna returns to the node at Brougham Castle and subsequently heads southeast to Smetriadum (159), the Roman fort at Bainbridge. The spelling of Smetriadum may be correct, dum meaning ‘summit of hill’, though since the name already includes the hill-letter it is perhaps more likely that dum was originally dunum. In any event the name refers to a location on the top of a high hill, a definition which is entirely appropriate for the Bainbridge fort (though of course the name will originally have been that of an Iron Age fort/settlement perhaps, but not necessarily, on the same site). Note that the river-letters corresponding to the hill-letters sm and r in Smetriadum are br and s. The b appears in the name Bain of the river flowing past the fort, and the r and s appear in the then name, Isur, of the river Ure, which is joined by the Bain immediately north of the fort (but see the entry for Smetriadum in the Alphabetical List, where it is suggested that the then name of the river Ure might actually have been Urus). After Smetriadum come Clindum (160), Carbantium (161) and Tadoriton (162) and these names have survived, at least in part, in the modern names Clint, Harrogate and Tadcaster. Note that these three places are aligned with one another in a northwest to southeast direction and are encountered in that order by someone coming down from Bainbridge. The modern village of Clint stands on a steep hill and the name Clindum comprises the old-style element nd meaning ‘hill summit’ qualified by the old-style element cl meaning ‘steep hill’. The Roman fort called Clindum may have been at the top of that steep hill (though of course whilst Iron Age Clindion/Clindonion will have been at the top of a steep hill the Roman fort Clindum need not have been on the same site). The initial har of Harrogate is merely an anglicised version of the car of Carbantium but the bant element cannot have been present in the original Celtic name – it must have been band meaning ‘high hill summit’. The meaning of Tadoriton is not clear, though this will have been the Celtic name of Tadcaster, Calcaria being the official Roman name. The meaning traditionally given to the name Tadoriton is 'grandfather ford', though this seems somewhat fanciful, unreal. It is more likely that the Celtic name had been somewhat of the form Cartadoriton or Cardadoriton, where cart means 'steep hill high' and card means 'steep hill summit', that some over-zealous medieval copyist mistook the initial car for an erroneous duplication of the initial part of the previous Ravenna name, Carbantium, and that he then deleted the car, thus leaving Tadoriton or Dadoriton, the latter then changing to Tadoriton. The dorit element of the name means 'summit of hill high'. The Celtic name will not have applied to the site of modern Tadcaster, but to an Iron Age fort/settlement up on the high ground to the west. The Romans may then have adopted the name of that fort/settlement and applied it to an early fort which they built on the low ground near the river Wharfe.


11     After Tadoriton Ravenna appears to follow an alignment from Tadcaster to Ribchester, for Maporiton (163) appears to have survived in Bramham (Mapor → Mabor →Bram), Loxa (165) in Exley Head, in the southern outskirts of Keighley, and Locatreve possibly in Trawden (trev → traw). For Maporiton Rivet and Smith suggest the meaning 'the young man's ford' or 'son's ford' (as opposed to the 'grandfather ford' at Tadoriton), though this seems fanciful. It is much more likely that Maporiton was originally Maboridon, where borid is the topographical element seen as bered in Bereda at Plumpton Wall in Cumbria. The element means 'high hill summit' and the Plumpton Wall fort was indeed built at the top of a high hill (though Iron Age Bereda need not have been on exactly the same site). It is thus most likely that Iron Age Maboridon was a fort or settlement up on top of one of the hills around Bramham and the name was transferred by the Romans in the Maporiton form to Bramham itself. Alitacenon (164) was thus between Bramham and Exley Head. It was probably the Roman fort to the north of Adel, which is roughly halfway between Bramham and Exley Head. Indeed it is possible that the very name Adel is derived from the initial Alit of AlitacenonLoxa at Exley Head comprises the inversion-type hill-letter l2 qualified by the old-style element ocsa meaning 'steep hill'. Note that the hill-letters s and l correspond to the river-letters b and t and these survive, with the b changed to v and then the latter anglicised to w, and with the t changed to th, in the river-name Worth, this river flowing north past Exley Head to join the Aire. The river-letter r in Worth indicates that the people who used the hill-letter m did at some point settle in the Worth Valley, though the hill-letter m does not appear in the name Loxa. The locat of Locatreve is an inversion-type element meaning 'hill steep high'. It is not clear whether the reve part of the name comprises the old-style element rev meaning 'hill slope' or is just the hill-letter r with an eve ending. The next two names are rather more difficult, so one can jump ahead to Uxela (169), Lucotion (170) and Corda (171). The first and third of these names appear to correspond to the Roman forts at Chesterfield and Leicester respectively, and Lucotion appears to correspond to Loughborough, which lies on a straight line connecting Chesterfield and Leicester. Uxela is a compound name referring to a location adjacent a steep hill, and there is indeed a steep hill in Chesterfield. Ekwall gives the form Lucteburne for Loughborough in 1086 and suggests a meaning ‘Luhhede’s burg’. However, since the form Lucteburne includes the l, the c and the t of Lucotion, and in exactly the same order, a derivation from the Roman name seems much more likely. Ravenna’s Corda might possibly be derived from the tribal name Coritani. On the other hand it might simply be an old-style topographical name meaning ‘steep hill summit’, just as the other name for Leicester, Rate, could be an inversion-type topographical name meaning ‘hill high’. It is of course also possible that Corda lost its initial C (as did Carmis (42) and Carduaravenatone (9)) and that orda simply turned into Rate. Returning to Ravenna’s Cambroianna (167) and Smetri (168), these appear to have been located between Trawden and Chesterfield. The name Cambroianna was probably Cambrolanda originally (cf. Ravenna’s Gabaglanda and the later Amboglanna of the Notitia Dignitatum), this name referring to a location on top of a steep, high hill. The obvious candidate is the fort at Slack, just west of Huddersfield. The existence of Cambrolanda at Slack might account for an error in Iter II of the Antonine Itinerary, for this Iter gives 20 Roman miles as the distance from Calcaria at Tadcaster to Cambodunum and then 18 Roman miles for the stretch from Cambodunum on to Manchester. The total mileage is of course far too low. But 20 Roman miles from Tadcaster takes us to Eccleshill in the northern outskirts of Bradford, and Eccleshill is located on top of a high, steep hill, which is what the name Cambodunum indicates. Moreover, the very name Eccleshill suggests that there was a church on the site when Anglo-Saxon settlers arrived in that area, and Bede’s Campoduno may well have been the same place as the Camboduno of the Antonine Itinerary. It thus appears sensible to identify Eccleshill as Camboduno/Campoduno. There is nonetheless an error in the name, since the b meaning ‘high’normally comes before the hill-letter. There must therefore have been another hill-letter after the b, perhaps an r, the name then having two old-style elements, cam and br, and the inversion-type element duno. There would then be on the road from Tadcaster to Manchester two different places with names starting with Cambro, and this may have confused some inattentive medieval copyist. He gave the distance of 20 Roman miles from Calcaria to Cambrodunum (Eccleshill) and then in the next line the distance of 18 Roman miles from Cambrolanda (Slack) to Manchester, but did not notice that he had overlooked the stretch from Cambrodunum to Cambrolanda.  Smetri is a bigger problem. It was presumably somewhere between Slack and Chesterfield, but no obvious candidate presents itself. The name may be spellt correctly, and indicates that the fort was adjacent a high hill (met = ‘hill high’), but the Celtic name may have been Smedri, referring to a location on the summit of a hill. Looking at the above group of names, Brocara to Corda, whilst one or two of the identifications may perhaps be thought a little shaky, the group as a whole is firmly anchored to known Roman forts at Brougham Castle, Bainbridge, Tadcaster, Ribchester, Chesterfield and Leicester. It thus seems safe to conclude that these Ravenna names were all located in the North of England and have no connection with the Selgovae or any other tribe living in what is now Scotland.


12     But having gone away down south to Leicester, the Ravenna Cosmography has to come back north again to Trimuntium at Newstead. Now, if one studies Ravenna’s names from Navione (106) to Corda (171), one will note that the only part of Northern England (south of Hadrian’s Wall) which Ravenna has not yet covered is the region to the east of the line joining Brough-on-Humber, Stamford Bridge, Thirsk and Dinsdale Park on the Tees. Praesidio is the Notitia Dignitatum name immediately after York, and the next Ravenna name after Corda (171) is Camulosessa Presidium (172,173), so it seems sensible to regard the ND Praesidio and Ravenna’s Camulosessa Presidium (172,173) as being one and the same place, and to identify Malton as that place. Camulosessa on its own will have been the name of an early fort located at the top of a steep slope on the north bank of the river Derwent. It is tempting to think that the name Presidium/Praesidio was applied to the later fort, traces of which can be seen today in the area known as Orchard Fields, the full name of the later fort being Ravenna’s Camulosessa Presidium. However, the name Camulosessa Presidium appears to have been taken from a first century map, so the fort overlooking the Derwent was probably already called Camulosessa Presidium, the reason for this not being entirely clear. But it may be that the fort at Malton enjoyed some special status as a forward command centre at the time when the 9th legion was still based in Lincoln but was active up the eastern side of the Pennines after Vespasian gave the command to bring the Brigantes under Roman control. Camulosessa stood on the river Derwent, previously the Derbentione. The tione part of this name is just the river-letter t applied by those who coined inversion-type names in the hill-letter l2, together with an ione ending. The Derb element comprises the river-letter t (changed to d, the t having been applied by those who used the hill-letter l1), r and b, so the corresponding hill-letters l, m and s, which are all present in Camulos, must be in the same chronological order within the name Camulos as the river-letters in Derb. It follows that the Celtic place-name must actually have been somewhat of the form Lucamosessa (where the hill-letter m and the element luc meaning 'hill steep' are used in an inversion-type manner) and the Lucamo of this form was simply changed by the Romans to the more familiar Camulo in Camuloduno, it being noted that the Romans had held Camuloduno at Colchester for some 20 years when they built their fort at Malton.  It is thus clear that the name Camulosessa has nothing to do with a Celtic war god called Camulos, as has been assumed in the past, and the river name Derventione has no connection whatsoever with oak trees, as has also been assumed in the past. 


13     The clue to the identification of the subsequent Ravenna names up to Celovion (178) lies in the fact that the forts at Chesters on the North Tyne, Corbridge and Ebchester are aligned with one another - and with Durham. This is hilly country, so the fact that these four locations are aligned with one another implies that the sites of the forts were selected, and the forts built, before the road builders arrived in that region. When, subsequently, the main north-south road, Dere Street, was built, it called at Ebchester and Corbridge, but not at Durham and Chesters. One can see, however, that Abisson (175) must have been at Durham, Ebio (176) at Ebchester, Coritiotar (177) at Corbridge and Celovion (178) at Chesters on the North Tyne. Ptolemy’s Epiacum was presumably the same place as Ravenna’s Ebio (176), the fort at Ebchester. The AI form of the name is Vindomora, which will have been Bindomora originally, this being a river-name of the kind comprising a river-suffix, here mora, attached to a place-name, here Bindo, meaning 'high hill summit'. Bindo will have been the name of the promontory fort on the summit of the high hill on the opposite side of the river Derwent from Ebchester. Bindomora will have been one of the two alternative names of the river Derwent, the other of course being Derbentione (see the entry for Ebio in the Alphabetical List). Ravenna's Ebio might be Bindo without its nd and with an E stuck on front, and Ptolemy's Epiacum might just be an attempt to write Ebio in a more classical Latin form.  Coritiotar (177), as noted, was a fort at Corbridge, probably the early fort at Red House. The Celtic name will have been Coritisotar (see the entry for Coritiotar in the Alphabetical List) and the later industrial/military complex at Corbridge took the name Coritisotaroppidum. This form, with omission of some internal letters (a common change in Romano-British place-names) became Corstopidum and this form, with the common d→t change, yielded the Corstopitum of the Antonine Itinerary. Celovion (178) appears to be the earliest form we have for the name of the fort at Chesters on the North Tyne. But Celovion will have been a Flavian fort and need not have been on exactly the same site as its Hadrianic successor. The name will have been transferred to the Flavian fort from the Iron Age settlement standing on the steep slope (Celov means ‘steep hill slope’) immediately east of the village of Wall, on the eastern side of the North Tyne. The name was transferred, slightly modified to Celunno, to a later Trajanic frontier fort built nearby, presumably on the Stanegate and possibly close to the point where the road crossed the North Tyne. Later still the name was transferred to the Hadrian’s Wall fort built a little upstream, at Chesters, and at some point the name changed to Cilurno. As to Abisson (175) at Durham, the ab part of the name is just the river-letter b, which is changed to v in the then name, Vedra, of the river Wear, the river on which Durham stands. Abisson will originally have been the name of the Celtic promontory fort now called Maiden Castle. The name will simply have been transferred to a Roman fort built nearby, probably on a site which would enable it to control a crossing of the Wear.


14     The Ravenna name between Camulosessa Presidium (172,173) and Abisson (175) is Brigomono (174). Now, if one draws a line on a map between Malton and Durham one will note that that line passes straight through a village called Great Broughton, and this may well be the site of Brigomono.  The elements of Brigomono appear to be the transitional element brig, meaning 'high hill steep', and the hill-letter m, the name referring to a location adjacent a steep, high hill, the hill in question being the escarpment just south of Great Broughton.


15     Having brought us up from Malton to Chesters, Ravenna proceeds to use Chesters as a node, for a first route out of Chesters goes via the village of Thockrington to Learchild and then on to Berwick-on-Tweed on Margary road 87. Thockrington will be the site of Ravenna’s Itucodon (179), the don element indicating that the Roman fort was located on the top of raised ground, as Thockrington is. The ituc element looks incomplete – it has the t meaning ‘high’ and the c meaning ‘steep’, but the hill-letter is missing. Perhaps it was an l, giving an element lituc, like the litac of Alitacenon  (164). Learchild was thus the Maromago (180) of Ravenna. The Learchild fort has in the past been considered to be the Alavna/Alauna of Ravenna and Ptolemy, the name having been transferred from the river Aln to the fort, but the fort appears rather too far from the Aln to have taken its name from that river. The mag element of Maromago refers to a location adjacent a steep hill, which is certainly true of Learchild, but the place-name was transferred to Learchild from Lacerocomaguve, the Iron Age promontory fort at Brinkburn Priory, within a loop of the river Coquet. If one rewrites the Celtic name as La [ce] ro [co] maguve, deletes the letters in brackets and replaces the uve ending with a simple o, one obtains the form Laromago. The l/m interchange then yields the name Maromago (cf. Celtic Lelamon → Ravenna’s Melamoni). After Maromago Ravenna takes us up to Duabsissis at Berwick-on-Tweed. This name is discussed in detail in Chapters 4 and 19.


16     The other route out of Chesters goes to a place called Venutio (182), which cannot be identified here but the venut part of the name may be a variation of vent, referring to a location on the side of a high hill. There is a side valley called Wainhope which joins the valley of the North Tyne to the north of Chesters, but whether the wain of Wainhope has any connection with Venutio is not clear. But Ravenna appears to use Venutio as a node, because the next-following names are Trimuntium (183) at Newstead, Eburocaslum (184) apparently at Broomholm and Bremenium (185) at High Rochester. Ravenna’s Alavna (187) and Ptolemy’s Alauna thus appear to have been at Alnwick, and Ravenna’s Coccuveda (186) will have been on the river Coquet at the point where a line joining High Rochester and Alnwick crosses the river, that is to say in the vicinity of Sharperton. Note that the caslum of Eburocaslum may be just a latinised version of the Celtic name or name-element uxelum/ucselum, this element cropping up at various points in the present study. The point is that elsewhere in Britain elements of that kind have given rise to river-names of the Esk, Exe, Axe, Usk family, and Broomholm is of course on the river Esk which flows into the Solway. The bur of Eburocaslum appears to be just an element in the hill-letter r, the element meaning ‘high hill’, and the ucs of ucselum refers to a ‘steep hill’. The Broomholm fort is indeed located at the top of a high, steep hill. But it is alternatively possible that Eburocaslum is a river-name transferred to the Roman fort, the name being of the kind comprising a river-element, here Ebur, a compound in the river-letters b and r, attached as a prefix to a land-name, here caslum.


17     When one looks at the above group of Ravenna names, Brocara to Alavna, one can see that a fairly high proportion of the names, including Clindum (Clint), Carbantium (Harrogate), Tadoriton (Tadcaster), Loxa (Exley Head), Locatreve (Trawden), Lucotion (Loughborough), Ebio (Ebchester), Coritiotar (Corbridge) and Itucodon (Thockrington) have survived at least in part in the modern names, so there should now be no doubt in anyone’s mind that apart from Trimuntium and Eburocaslum all of those places were in fact in the north of England.


18     The first two maps below show the locations of most of the remaining place-names in the north of England in the Ravenna Cosmography. The second map only goes as far south as Uxela at Chesterfield, so the third map is presented merely to show the relationship of Lucotion (170) and Corda (171) to Uxela.  Again the names and their numbering are those provided by Richmond and Crawford. The identifications are as follows:

156 Brocara                      Brougham Castle

157 Croucingo                  Croglin

158 Stodoion                    Nether Denton

159 Smetriadum               Bainbridge

160 Clindum                     Clint

161 Carbantium                Harrogate

162 Tadoriton                   Tadcaster

163 Maporiton                  Bramham

164 Alitacenon                  probably Adel

165 Loxa                           Exley Head, Keighley

166 Locatreve                   Trawden

167 Cambroianna              Slack, west of Huddersfield

169 Uxela                          Chesterfield

170 Lucotion                     Loughborough

171 Corda                          Leicester

172,173 Camulosessa Presidium             Malton

174 Brigomono                Great Broughton

175 Abisson                     Durham

176 Ebio                           Ebchester

177 Coritiotar                   Corbridge

178 Celovion                    Chesters on the North Tyne

179 Itucodon                   Thockrington

180 Maromago                 Learchild

181 Duabsissis                 Berwick-on-Tweed

182 Venutio                     unknown

183 Trimuntium               Newstead

184 Eburocaslum             Broomholm

185 Bremenium               High Rochester

186 Coccuveda                 in the vicinity of Sharperton

187 Alavna                       apparently Alnwick



 [This page was last modified on 02 April 2022 ]