[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 1: Navione to Fanococidi


1     We start a study of the Roman place-names in northern England at Navione (106), at Brough-on-Noe, the modern river-name Noe clearly being derived from Navione.  Navione is used as a minor node in Ravenna, for the next name is Aquis Arnemeze (107) at Buxton, which is to the southwest of Brough-on-Noe, whereas the next again is Zerdotalia (108) at Melandra Castle, to the northwest of Brough-on-Noe. The initial Ze of Zerdotalia is an error, being a carryover from the preceding name, Arnemeze. The correct original spelling of Zerdotalia may have been Cerdodalia, a compound in the old-style element cerd (meaning ‘steep hill summit’) and the inversion-type element dal (an element of the duno-type but using the hill-letter l), the chronological order of the hill-letters thus being first r and then l. The corresponding river-letters are s and t, and these appear in exactly this order in Ptolemy’s river Seteia. Now, three rivers come together to form the river Mersey and one of them is the river Etherow, which flows past Melandra Castle. It would thus appear that the river Etherow was regarded as the main river in Roman days, the name Seteia thus being applied to the combination of the Etherow and the Mersey, so that the stretch of water now called the Mersey estuary was indeed the Seteia estuary of Ptolemy. Zerdotalia was also used as a minor node, for the next-following name in Ravenna is Mantio (109) at Manchester, to the northwest of Melandra Castle, whereas the next-again name is Alicuna (110) at Castleshaw, to the north of Melandra Castle. Note that Mantio was probably originally Mandio, the mand element referring to a location on the top of raised ground. Presumably the Manchester fort was abandoned at some point and later re-commissioned under its new AI name, Mamucio, thought by scholars to mean ‘breast-shaped hill’.  Alicuna at Castleshaw lay on a straight line between Zerdotalia at Melandra Castle and Camulodono (111) at Skipton. The fort called Camulodono presumably stood up on the high ground now partly occupied by Skipton castle, the cam element in the name referring to the steep drop down to the Eller Beck, a tributary which joins the river Aire just south of Skipton. Note that the l of Eller is the river-letter l applied to minor rivers by those who used the hill-letter n, present in the dono element of Camulodono, and the r of Eller is the river-letter corresponding to the hill-letter m in Camulodono. Note, too, that the similarly-named Camboduno at Kempten in Bavaria, also including the hill-letters m and n, stood at the top of a steep escarpment on the east side of the river Iller, which is of course the same name as Eller. But even without this correspondence between land-name and river-name one can see that Camulodono was at Skipton simply by looking at a map, since Ravenna used Skipton as a major node. The names immediately following Camulodono (111) up to Derventione (122) lay to the northwest of Skipton, the initial ones of those names lying on Margary road 722. Then comes Ravonia (123), presumably originally Bravonia, at Burwen Castle, Elslack, to the west-southwest of Skipton, this lying between Skipton and the next-following name, Bresnetenaci Veteranorum (124) at Ribchester. Then comes Pampocalia (125) at Ilkley, to the southeast of Skipton, this lying between Skipton and Lagentium (126) at Castleford.

2     Before discussing individual ones of the names between Camulodono and Lagentium it is helpful to look briefly at Ptolemy’s place-names in the territory of the Brigantes. These are EpiacumVinoviumCaturactoniumCalagumIsurium,  RigodunumOlicanaEboracum, and CamunlodunumEpiacum will have been the same name as Ravenna’s Ebio, which was at Ebchester, this being discussed in detail below, and VinoviumCaturactoniumIsurium and Eboracum were of course respectively at Binchester, Catterick, Aldborough and York. Ptolemy’s Camunlodunum will have been the same place as Ravenna’s Camulodono, which was at Skipton, as discussed above. And if Ptolemy’s Olicana was the same place as Ravenna’s Alicuna, then it was at Castleshaw. Rigodunum comes between Isurium at Aldborough and Olicana at Castleshaw in Ptolemy’s list, and if one draws a line on a map between Aldborough and Castleshaw, one will note that that line passes through a village called North Rigton, to the southwest of Harrogate. This village appears from the OS map to lie up near the top of a steep hill. Indeed the village appears from the map to extend onto the top of the hill, which is where one would expect to find a place called Rigodunum, the rig element being an inversion-type place-name element meaning ‘hill steep’ and dunum referring to the summit of a hill. It does therefore appear that North Rigton should be identified as Ptolemy’s Rigodunum.  It appears, then, that in listing Epiacum to Olicana Ptolemy was simply following an itinerary starting at Ebchester in the north and ending at Castleshaw in the south. His Calagum thus lay somewhere between Catterick and Aldborough. It may have been on the eastern side of the Pennines, but then again it may have been the same place as the Galacum of Iter X of the Antonine Itinerary, and that was west of the Pennines. And it may be that that place, Calagum/Galacum, was the same place as Ravenna’s Caluvio, which was somewhere to the northwest of Skipton on Margary road 722. Calagum/Galacum/Caluvio is thus most likely to have been at Ingleton, at the junction of Margary roads 73 and 722. The identification appears correct since the distance by road between Ingleton and Ribchester appears to correspond to the mileage given in Iter X for the journey between Galacum and Bremetonaci at Ribchester. Ptolemy’s traveller will then have come over from Catterick to Bainbridge (perhaps there was actually a Roman road between those two places), where he would have got onto Margary road 73 for Ingleton. He would have come back east again by taking road 722 from Ingleton to Skipton and will then have gone on to Ilkley where he would have turned left onto Margary road 720 for Aldborough. But note that this means that no matter who placed the names Epiacum to Olicana in Ptolemy’s Geography, whether Ptolemy himself or a much later copyist, he was apparently unaware of the relationship of Camunlodunum to the names from Epiacum to Olicana, for if he had known where all of these places were, then surely he would have placed Camunlodunum between Calagum and Isurium in his list. It thus seems more sensible to leave Ptolemy’s Calagum east of the Pennines, a Roman post somewhere between Catterick and Aldborough, most probably the Roman fort/settlement at Healam Bridge, on Dere Street.


3     We can now go back to consider the Ravenna names after Camulodono at Skipton. The first place is Caluvio at Ingleton, where the Caluv element means ‘steep hill slope’. The name will have been transferred to Ingleton from the Iron Age settlement known as Yarlsber Camp (NGR: SD 710 726). The settlement is about one kilometre east of Ingleton and does indeed stand on a steep slope, the highest part of the settlement being around the 250 metre contour and the lowest part at around 235 metres. The next three names in Ravenna are GalluvioMedibogdo and Cantiventi.  

3.1     Cantiventi should be identified as the Roman fort at Watercrook, just south of Kendal. The name appears to be a river-name of the kind comprising a river-suffix attached to a place-name including one or more hill-letters. The river-suffix would be vent (seen also in Derventio and Aventio) comprising the river-letters b (changed to v) and t corresponding to the hill-letters s and l  respectively. The place-name may have been the transitional element Cant (meaning ‘steep hill high’) but is perhaps more likely to have been the old-style element Cand (meaning ‘steep hill summit’). Candion will have been the name of the hillfort now known as Castlesteads, The Helm, overlooking the river Kent from the summit of a steep hill a little south of Oxenholme. The Roman fort at Watercrook lies in a loop of the river Kent, so the Romans will simply have transferred the name of the river to their fort, Candiventi changing at some point to Cantiventi.

3.2     Now, if one draws a line on a map between Ingleton and Watercrook, one will note that that line crosses the river Lune at Casterton, so it would appear that Casterton should be identified as Galluvio. The Celtic name will have been Gabluvion, where Gabluv means ‘steep high hill slope’. Gabluvion will have been the name of the hillslope fort known as Castle Hill (NGR: SD 650 779) to the southeast of Casterton, the name simply having been transferred to Casterton by the Romans with the change blll. The Romans will then have transferred the name Galluvio to the river, only the tail-end of the name surviving in the modern river-name Lune, of course with the common change of v to n.  The distance from Watercrook to Lancaster corresponds to the Iter X mileage between Cantiventi (using Ravenna’s form of the name) and Galava, so Lancaster should be identified as Galava, the modern name being derived from the lava part of the Roman name, again with the common change from v to n. Again the gal element refers to a steep slope. And the distance between Lancaster and Burrow-in-Lonsdale corresponds to the Iter X mileage between Galava and Alone, so the fort at Burrow-in-Lonsdale should be identified as the Antonine Itinerary Alone. These identifications might explain the little mystery of the three different vowels after the initial L in the river-name Lune, the valley-name Lonsdale and the city-name Lancaster. The three modern names are simply derived from different Roman names. The river-name Lune is derived from the final element of Galluvio, with the v changed to n. The valley name is derived from Alone at Burrow-in-Lonsdale, and the city name Lancaster is derived from the final part of Galava, again with the change from v to n. Note that there is still an error in the Antonine Itinerary since Iter X gives the distance from Alone to Calacum as 19 Roman miles, whereas the distance from Burrow-in-Lonsdale to Ingleton is only about 9 Roman miles, assuming the traveller first had to get from Burrow-in-Lonsdale to Margary road 7c and then traveled south on that road to somewhere in the vicinity of Lower Bentham, and was there able to turn left on to a road to Ingleton. What this means is that there is an intrusive x in the mileage in Iter X, though since this particular error occurs in other places in the Antonine Itinerary it is not an error over which one should lose any sleep.

3.3      Ravenna’s Medibogdo (114) will have been a Roman post built in the vicinity of the Iron Age hillfort at Kitridding (NGR: SD 583 843), standing more or less on a straight line between Casterton and Watercrook. The Celtic name of that hillfort was most probably of the form Medibogldo or Mediboglodono (see the entry for Medibogdo in the Alphabetical List), this then being transferred, with minor amendment, to the nearby Roman post.

3.4     The name after Cantiventi in Ravenna is Iuliocenon (116) and, whilst no suggestion is made here as to the meaning of the name (but see the entry for Iuliocenon in the Alphabetical List), it was evidently the name of the Roman fort at Ambleside. After Iuliocenon comes Gabrocentio (117), the fort at Hard Knott pass, gabr being an old-style element in the hill-letter r and meaning ‘steep high hill’. Cent is a transitional element in the hill-letter n, the meaning being exactly the same, but literally it is ‘steep hill high’. The gabr element developed, after the arrival of the English language in that area, into the Hard element of the modern name Hard Knott, whereas cent developed into Knott. The name after Gabrocentio is Alavna (118), which must have been the name of the Roman fort at Ravenglass, near the mouth of the river Esk. But Alavna is a river-name which has been adopted by the Romans and applied by them to a fort which they built on the banks of the river.  In other words, the Cumbrian river now called the Esk was in fact called the Alavna when the Romans arrived in that part of England. Perhaps the fort called Alavna was abandoned at some time and was later re-commissioned under the land-name Isca (meaning ‘hill steep’), this name later being transferred by the Romans to the river, now the Esk. It is not of course surprising that there should be a place called Isca on the banks of a river called the Alavna, since Alavna, as explained in Chapter 5, is a compound river-name in the river-letters l and b, where the b has changed to v, and b is of course the river-letter corresponding to the hill-letter s in Isca.

3.5     The next name in Ravenna is Bribra (119), which was the fort at Moresby. Bribra looks very odd but is in fact a name with two elements both of the form br. One br is a land-name element in the hill-letter r and meaning ‘high hill’, whereas the other br is a compound river-name in the river-letters b and r, the whole name thus actually being that of the stream now called the Lowca Beck, the river-name Bribra then being transferred by the Romans to the fort which they built up on the high ground immediately to the south of the stream. After Bribra comes Maio (120) which, although no suggestion is made here as to the meaning of the name (but see the Alphabetical List), was evidently the name of the fort at Burrow Walls in Workington. And this name is the reason why Maia at Kirkbride has the adjunct ‘Fanococidi’ – it was necessary to distinguish between Maia and Maio located fairly close to one another on the Cumbrian coast (but see Chapter 20: Rome's frontiers in northern England, where it is argued that Fanococidi was a Flavian fort at Kirkbride and Maia a Trajanic fort up on Fingland ridge). After Maio in Ravenna comes Olerica (121) and that was the name of the Roman fort at Maryport. The name, however, does not look quite right. There is no problem with the l and r – they are just two different hill-letters – but the ca at the end looks very odd. It was probably originally isca, referring to the steep hill up to the fort from the south or the steep slope, cliff actually, on the western side of the fort. The name was thus most probably Olerisca, and Olerisca is one of the Romano-British names which have been reversed at some stage, for if one reverses Olerisca and adds a dunum to the end one obtains Acsirelodunum. If one then deletes ir, because it makes the name unwieldy and difficult to pronounce, what is left is Acselodunum, which is of course the Axelodunum of the Notitia Dignitatum.  The Notitia indicates that the Cohors I Hispanorum was stationed at Axelodunum and various inscriptions referring to that unit have been found at Maryport. There can thus be no doubt whatsoever that Axelodunum is in its correct position in the Notitia list and is not Uxellodunum (Castlesteads) in the wrong position (and wrongly spellt), as suggested by various scholars in the past. Name reversal is not peculiar to Olerisca. Ptolemy’s river-name Isca was reversed and a grammatical um added on to the end of the reversed name to yield Acsium, which is the river Axium of Ravenna, both Isca and Axium referring to the river Exe in Devon. In addition the Ucsel of Uxelis was reversed to yield Lescu and this has become the Lisk of modern Liskeard. Furthermore, Tagea has been reversed to yield the geath of modern Strageath and Decha has been reversed to yield the keith of modern Inverkeithing. In addition Ravenna’s river-name Novitia has been reversed and the v dropped to yield the river-name Tyne (East Lothian). After Olerica in Ravenna comes Derventione (122) at Papcastle. This name is discussed in detail in Chapter 19: The rivers of Roman Britain.


3.6     To get to the next place listed in Ravenna, Ravonia (123), it is necessary to go back to the node, Camulodono at Skipton, and then head west-southwest on Margary road 72a to Burwen Castle, Elslack. The Celtic name was probably Bravonia, and the modern name Burwen may indeed be derived from that form. The Celtic name most probably includes the element brav, which is a banva-type name-element, but using the hill-letter r instead of n. The element indicates that the fort was built on the slope of a high hill. After Bravonia comes Bresnetenaci Veteranorum (124) at Ribchester. Ribchester stands on the river Ribble, which is thought to be the Belisama of Ptolemy. But Belisama cannot be the Celtic name of the river since it includes the river-letter l, which refers to minor rivers, as well as the river-letter m, which refers to major rivers, so one cannot have both river-letters in the name of one and the same river. The Celtic name of the river was probably similar to Belisama and was simply changed to Belisama by Gaulish troops who were stationed at Ribchester and were familiar with the name of the goddess Belisama, who was worshipped in Gaul. The Celtic name probably began with Be and ended with sama, which would be consistent with the use of the river-letter m for a major river. And since the AI and ND forms of the name of the Ribchester fort both include the letter m, it is reasonable to assume that the m is the hill-letter m corresponding to the river-letter r. It would thus appear that the Celtic name of the river was actually Berisama. The hill-letters corresponding to the river-letters brs and m in Berisama are smr and n, so one can see that Ravenna’s Bresnetenaci is probably fairly close to the Celtic original. But the hill-letters smr and n must occur in the same chronological order within the place-name as the river-letters brs and m in the river-name. This seems to require that the Celtic place-name was of the form Vresmedenaci where Vr means ‘side of hill’ or ‘slope’ and esm is an old-style compound in the hill-letters s and m, the element Vresm thus having the hill-letters rs and m, but in the chronological order smr corresponding to the order of the river-letters br and s in Berisama. The denac element is a duno-type ending with the hill-letter n qualified by c meaning ‘steep’. The hill-letter in duno-type endings is not normally itself qualified, but in Omirededertis (25) the dert element means ‘summit of hill high’, so there seems no good reason to object to an element denac meaning ‘summit of hill steep’. It thus seems clear that the Celtic name of the Ribchester fort was Vresmedenaci, but that by the date of preparation of the first century map used by the compiler of Ravenna the v had changed to b and the d to t (in themselves common changes in Romano-British place-names) so that the name actually on the map was Bresmetenaci. The first n in Ravenna’s Bresnetenaci is probably just the result of m/n confusion at some stage of medieval copying. The form in Iter X of the Antonine Itinerary is Bremetonaci, which is not so very far from Bresmetenaci – it has dropped the s and there is an unimportant change of vowel after the t. The ND form is Bremetenraco which, like the AI form, has dropped the s and furthermore has inserted an r after the n. This presumably indicates that the part of the Notitia Dignitatum which includes Bremetenraco is later in date than Iter X of the Antonine Itinerary. But note that the name Vresmedenaci is not actually appropriate for the location of the Ribchester fort. The name will have been that of the hillfort known as Portfield Camp (at SD 746 355, to the southeast of Whalley) and was simply transferred to Ribchester by the Romans. The name Vresmedenaci is unusual in including both a v meaning ‘slope’ and a d meaning ‘summit’, but the hillfort is also unusual in standing on a site which apparently gives the impression of having been artificially levelled (see, for example, the entry for Portfield Camp on the Pastscape website of Historic England), that site being at the top of a steep slope. The location at the top of a steep slope explains the denac element of the place-name, and if one ignores the artificial levelling the land around the site rises from the southeast to the northwest. It may also rise from the southwest to the northeast in the direction of the higher ground to the northeast of the hillfort. The Vr element of the name is thus also appropriate. The name Vresmedenaci is thus entirely appropriate for the hillfort. What this means is that the nearby river Calder must have been the river Berisama and since Ptolemy refers to the Belisama estuary it is clear that the name Berisama was also applied to that part of the modern river Ribble downstream of the confluence of the two rivers. The Roman fort at Ribchester thus actually stood on the banks of the river Berisama, hence the suggestion made above that the river-name was changed to Belisama by Gaulish troops stationed at that fort.


3.7     To reach the next Ravenna name, Pampocalia (125), one must go back to the node at Camulodono at Skipton and then head southeast to Ilkley. Pampocalia will originally have been Bamvocalia, the bamv element being a banva-type name but using the hill-letter m instead of n, the element bamv referring to a location on the side, the slope of a high hill. The cal element refers to a location adjacent a steep hill. The name Bamvocalia is thus a precise and correct topographical description of the location of the Ilkley fort. After Pampocalia in Ravenna comes Lagentium (126) at Castleford. The lag element means ‘hill steep’ and ent means ‘hill high’, the name as a whole thus referring to the high, steep hill immediately SE of the location of the fort. But note that the Celtic name may actually have been Lagendion (see the entry for Lagentium in the Alphabetical List).


4     After Lagentium Ravenna jumps to Valteris, an early form of Verteris at Brough-under-Stainmore. The next name, Bereda (128), at Plumpton Wall, indicates that the fort was built on top of a high hill, as indeed it was. Ravenna now moves on to Lagubalium (129) at Carlisle, the element lag being an inversion-type element meaning ‘hill steep’ and being qualified by the old-style element bal which means ‘high hill’. Note, too, that the river-letter corresponding to the hill-letter l in Lagubalium is t, and this appears, changed to d, in the name Eden of the river at Carlisle. Ravenna now proceeds to Magnis (130) at Carvoran which is used as a minor node, since the next name is Gabaglanda (131) at Birdoswald to the west-northwest of Carvoran, whereas the next-again name is Vindolande (132) at Chesterholm, to the east of Carvoran. Gabaglanda shows the c/g confusion which is common in Romano-British place-names and also the medieval omission of the letter m. The Celtic name will have been Cambaglanda, comprising three topographical elements, namely cam referring to a steep hill, bagl referring to a high, steep hill, and and referring to a hill summit. The name, an old-style name in the three hill-letters ml and n, thus refers to the location of the fort at the top of the steep, high escarpment on the north bank of the river Irthing. Note that the river-letters corresponding to the hill-letters m and l of Cambaglanda are r and t, and both of these are present in the river name Irthing, though with the t changed to thVindolande will originally have been Bindolande, where the Bind element, using the hill-letter n1, means  ‘high hill summit‘ and the and element, meaning ‘hill summit’ uses  n2. Such a Bind element normally refers to an Iron Age hillfort, so Bindolande will have been the name of the hillfort high up on Barcombe Hill, just east of Chesterholm. The Romans simply transferred the name to the Flavian fort which they built at Chesterholm itself.


5     Ravenna now starts a journey south to Decuaria (138) (also known as Petuaria) at Brough-on-Humber via Lincovigla (Lanchester), Vinovia (Binchester), Lavaris (Bowes), Cactabactonion (Catterick) and Eburacum (York). These are all well-known, so there is no need to take up too much space discussing them here. But there are three interesting points worth noting. Firstly, Lanchester is Longovico in the Notitia Dignitatum, this name presumably being a very slight modification and simplification of Lincoviglavicus, rather suggesting that an early fort there had been abandoned at some point, though its vicus lived on. Lincoviglavicus may have changed to Lincovicus and the latter form then changed to LongovicoLincovigla will have been Lincobigla originally, bigl being an old-style element meaning ‘high steep hill’. The area was later taken over by people who coined inversion-type names in the hill-letter n, the place being, for them, hill steep (inc) called bigla, written down as Incobigla. Then those who coined inversion-type names in the hill-letter l arrived on the scene and for them the place was a hill (l) called Incobigla, written down as Lincobigla. In Lincobigla the hill-letter l of the bigl element is l1, whereas the initial L of the name is l2. Secondly, Vinovia will have been Celtic Binovia – the fort stood up on high ground adjacent the river Wear. Thirdly, the name Cactabactonion is clearly not original, though evidently closer to the Celtic name than is the Caturactonium of Ptolemy. The Roman centre at Catterick is some four miles from the nearest waterfall, so the original name of the place can have had nothing to do with waterfalls! However, there is raised ground in the immediate vicinity, close to Thornbrough farm on the west side of the A1, so it is probable that the name is topographical. The Celtic form of the name was probably Castaractonion, where cast is a transitional element in the hill-letter s (meaning ‘steep hill high’ and similar to the cant of Cantiventi (115)), and ract an inversion-type element in the hill-letter r (meaning ‘hill steep high’ and similar to the arcot of Marcotaxon (225)). Note that the river-letters corresponding to the hill-letters s and r in Castaractonion are b and s, both of which are present in the name of the river at Catterick, the Swale, where the b has changed to v and the v has been anglicised to w. The l of Swale is probably just the river-letter l applied to minor rivers by those who used the hill-letter n, it being noted that the Swale is a tributary of the Ure. The change from the cast of Castaractonion to the cact of Cactabactonion is not without parallel – Lectoceto at Towcester will originally have been Lectoseto. But the change from the ract of Castaractonion to the bact of Cactabactonion is rather more difficult to explain. However, there are names where an r changed to v – for example the Brocara of Ravenna changed to the later Brocavo of the AI (i.e. later in date than the first century map used by the compiler of Ravenna) – and the further change from v to b is fairly common in Romano-British place-names, so this may account for the b in Cactabactonion. It is to be noted that if the s of Castaractonion is dropped what is left is Cataractonion, so now one can see the origin of the Caturactonium of Ptolemy and the Cataractoni and Cataractone of the Antonine Itinerary. These Ptolemy and AI names have nothing whatsoever to do with waterfalls or rapids, as is commonly supposed – they are just minor modifications of the topographical name Castaractonion.


6     After Decuaria/Petuaria Ravenna takes us back up north, calling first at Devovicia (139) at Stamford Bridge. Devovicia is simply De [r]v [entiovicus, with the text in brackets dropped and Devovicus modified slightly to Devovicia. Again the use of the term vicus suggests that the army had moved on by the time the place came to be known as Devovicia. The Antonine Itinerary indicates that Derventio was 7 Roman miles from York, this pointing to a location at Stamford Bridge. Ravenna now takes us from Stamford Bridge to Coganges (141) at Chester-le-Street via Dixiolugunduno (140). It would appear that we are following Margary road 80a and that Dixiolugunduno does in fact comprise two quite separate names Dixio and LugundunoDixio, presumably the Dicti of the Notitia Dignitatum, may have been at Deighton and Lugunduno at Dinsdale Park, where Margary road 80a crosses the Tees. Note that Dixio/Dicti may originally have been somewhat of the form Descoti. If the s is dropped one has Decoti, which may have changed to the Dicti of the ND. If, on the other hand, the intervocalic t is dropped one has Descoi, which might possibly have become the Dixio of Ravenna. But Descoti is an inversion-type Celtic name meaning 'summit of hill steep high', which is clearly inappropriate for Deighton. But the name would be entirely appropriate for the Celtic hill-fort at Roulston Scar. It may thus be the case that Descoti was the name of that hill-fort and the name was simply transferred by the Romans to a fort or posting station built nearby on Margary road 80a, most probably at or near Thirsk. But a duno-type element in the hill-letter s is extremely unusual - indeed the writer can bring only one example to mind, namely Descecanglion, which yielded the tribal name Deceangli on omission of s and the second c. It may thus be that the hill-fort at Roulston Scar had a name somewhat of the form Deriscoti, in which the inversion-type element Der, meaning 'summit of hill', is qualified by the original name of the hill-fort, namely Iscoti, this being an inversion-type name meaning 'hill steep high'. Duno-type elements normally come at the end of a name, but there are exceptions, such as Demerosesa and DolocindoDeriscoti may be another. What this would mean is that the hill-fort was built by people who used the hill-letter s and was then taken over by people who used the hill-letter r. This is exactly what happened in the case of Castaractonion at Catterick, only some eighteen miles distant. The earlier element in that name is the transitional element Cast meaning 'steep hill high' and the later element is the inversion-type element ract meaning 'hill steep high'. The form Deriscoti is thus entirely plausible and the modern name Thirsk may well be derived from it. Lugunduno at Dinsdale Park will have been at the top of the steep slope on the north bank of the Tees, just as Lugdunum at Lyon was at the top of the steep slope on the west bank of the river Saône. The un in the middle of Lugunduno may be intrusive or may just be the surviving ending of the original name, Luguno or Lugunum, the duno element being a later addition. Note that the name Lugunduno, or at least the duno element of it, was transferred by the Romans to the river, so that the bay at the mouth of the Tees was the Dunum bay listed by Ptolemy. Note, also, that the hill-letter l in Lugunduno corresponds to the river-letter t in the river-name Tees. Thus whilst, so far as the present writer is aware, no traces of a Roman fort have ever been found at Dinsdale Park, the place-name Lugunduno, it's position within Ravenna's list and the river-name Tees all point to there having been a Roman fort at Dinsdale Park, presumably an early fort built to guard or control the crossing of the Tees. Coganges is an odd-looking name, long thought impenetrable, but the building-blocks explained in Chapter 1 permit clarification. The point to note is that Chester-le-Street lies in a steep-sided valley, and the name Coganges includes the adjectival c/g in three places. The central element gan and final element ges are both old-style elements meaning ‘steep hill’. Now, the then name of the river Wear was Vedra, a river-name in the river-letters b (changed to v), t (changed to d) and r, these corresponding to the hill-letters sl and m, so it is possible that two of those hill-letters should appear in the name Coganges (in addition to the hill-letter n which is present in the gan element). The s is already present in the element ges, so it is possible that the first element of the place-name includes either an l or an mVedra is the same river-name as Bdora/Boderia (the Forth) and in both names r is the ‘youngest’ river-letter in the name, t changed to d the next youngest and b changed to v the oldest (this is explained in Chapter 19: The Rivers of Roman Britain). It follows that the hill-letter l (corresponding to the river-letter t changed to d) or m (corresponding to the river-letter r) in the place-name would have to be younger than the hill-letter s (corresponding to the river-letter b changed to v). This appears to require that the first element of the name Coganges must have been an inversion-type element. In other words the Celtic name of the fort at Chester-le-Street was either Lucoganges or Mucoganges. All that has happened is that the initial Lu or Mu has been lost or dropped at some stage to leave the Coganges of Ravenna. Note that those people who used the hill-letter n did not leave their mark on the river-name – there is no river-letter m or l in Vedra. This is not a problem – where there are several settlements on the banks of one particular river one cannot expect 100% correspondence between the river-letters in the river-name and the hill-letters in each and every one of the place-names. After Coganges comes Corielopocarium (142). It has been suggested in the past that this name refers to Corbridge, but Corbridge is mentioned later in Ravenna as Coritiotar (177) and Ravenna is about to begin a list of forts from Wallsend to the Solway, so it seems much more sensible to identify South Shields as Corielopocarium


7     The two maps below show the place-names discussed above and taken from a Flavian map of the north of England by the compiler of the Ravenna Cosmography. The names and their numbering are those provided by Richmond and Crawford. The identifications are as follows:


106 Navione                     Brough-on-Noe

107 Aquis Arnemeze         Buxton

108 Zerdotalia                  Melandra Castle

109 Mantio                       Manchester

110 Alicuna                      Castleshaw

111 Camulodono              Skipton

112 Caluvio                      Ingleton

113 Galluvio                     Casterton

114 Medibogdo                hill-fort south-southwest of Oxenholme

115 Cantiventi                  Watercrook

116 Iuliocenon                 Ambleside

117 Gabrocentio               Hard Knott

118 Alavna                       Ravenglass

119 Bribra                        Moresby

120 Maio                          Workington

121 Olerica                       Maryport

122 Dereventione             Papcastle

123 Ravonia                     Elslack

124 Bresnetenaci Veteranorum     Ribchester

125 Pampocalia               Ilkley

126 Lagentium                Castleford

127 Valteris                     Brough-under-Stainmore

128 Bereda                      Plumpton Wall

129 Lagubalium              Carlisle

130 Magnis                     Carvoran

131 Gabaglanda              Birdoswald

132 Vindolande               Chesterholm

133 Lincovigla                 Lanchester

134 Vinovia                     Binchester

135 Lavaris                      Bowes

136 Cactabactonion         Catterick

137 Eburacum                  York

138 Decuaria                    Brough-on-Humber

139 Devovicia                  Stamford Bridge

140 Dixio                         Roulston Scar/Thirsk

140 Lugunduno               Dinsdale Park

141 Coganges                  Chester-le-Street

142 Corielopocarium       South Shields


(Note that Corielopo and Carium are incorrectly shown as two separate places on the first map below. Corielopocarium is one name, referring most probably to South Shields. Note further that Valteris at Brough-under-Stainmore is not shown on the maps, and that Dixio at Roulston Scar/Thirsk is not shown separate from Lugunduno (140)). 



8     After Corielopocarium Ravenna gives the names from Serduno (143) to Maia Fanococidi (154,155), these names being discussed in detail in Chapter 8: the Trajanic frontier between Tyne and Solway.



[The text of chapter 15 is continued in Part 2: Brocara to Alavna. Part 2 is accessible by clicking here]


[This page was last modified on 02 April 2022]