[NB. Place-name identifications indicated below are explained in Chapters 10-16, which deal with Romano-British place-names in different regions of Britain.]
Chapter 19
The rivers of Roman Britain
1 As explained in Chapter 1 the different waves of Celtic settlers who arrived in Britain from the Continent used different letters to signify ‘hill’ and also different letters to signify ‘river’. Those who used the hill-letter r used the river-letter s, those who used the hill-letter m used the river-letter r, those who used the hill-letter l used the river-letter t and those who used the hill-letter s used the river-letter b. Those who used the hill-letter n make our lives a little more complicated since they used two river-letters – m for major rivers and l for minor rivers, where a minor river may be a relatively small river which flows directly into the sea or may be a tributary or headwater of another river. The normal chronological order of the hill-letters appears to have been n1, s, m, r, l1, n2, l2, though there is some regional variation. The hill-letter l2 was used by Celts who may have come to Britain sometime after about 150BC. The hill-letters n1 to n2 were then used by different groups of Celts who came to Britain at different times prior to 150BC.
2 It is important to note that a river-letter just means ‘river’. And just as compound place-names exist comprising two or more hill-letters, so compound river-names exist comprising two or more river-letters. For example, the river-name Abona comprises just one river-letter, b, with an ona ending. The compound river-name Alavna comprises the river-letters l and b, changed to v, so that the name Alavna actually means the ‘river Avna’, or, more precisely, the ‘river Abona’.
3 In compound old-style place-names the hill-letters occur in chronological order going from left to right, and in associated river-names the corresponding river-letters appear in exactly the same order. Thus, for example, the hill-letters in the old-style part Conionemedo of the name Elconionemedo (Launceston) are n and m. The corresponding river-letters m and r appear in exactly the same order going from left to right in the river-name Tamaris, the modern Tamar. The hill-letters in Londinium are l and n, the corresponding river-letters being t and m, these appearing in the same order in the river-name Tamesis. And in the old-style name Cambaglanda (Birdoswald) the hill-letters m and l correspond to the river-letters r and t, these appearing in exactly the same order, with the t changed to th, in Irthing, the name of the river which flows past the fort.
4 When hill-letters are used in an inversion-type manner the corresponding river-letters are in general used in the same way. For example in the name Elconionemedo, referred to above, the initial hill-letter l is used in an inversion-type manner - it appears at the beginning of the place-name qualified by the earlier place-name Conionemedo. The corresponding river-letter t is used in exactly the same manner - it appears at the beginning of the river-name, qualified by the earlier name of the river. In addition, the chronological order of the hill-letters in Marcotacson (Ravenna's Marcotaxon) is first s, then r and finally m, the hill-letters r and m being used in an inversion-type manner. The hill-letters thus appear in inverse chronological order - first m, then r, finally s - going from left to right within the place-name. The corresponding river-letters r, s and b also appear in inverse chronological order going from left to right within the river-name, the Ruthven Water. Note that the river-letter b changed to v and the s changed to th (presumably at some date later than the Roman period). Likewise the chronological order of the hill-letters in Lectoseto (Towcester) is first s and then l, the l being used in an inversion-type manner, so that the hill-letters appear in inverse chronological order going from left to right within the place-name. The corresponding river-letters also appear in inverse chronological order going from left to right within the river-name Tove, i.e. first t and then b, changed to v. It may be noted in passing that this demonstrates that the river-name Tove is Celtic and has no connection with any Scandinavian lady called Tofi, as is commonly supposed. Finally, the Celtic form of Ravenna's Ugrulentum was actually Rugulentum. Those who used the hill-letter n applied their river-letter l (for minor rivers) to that stretch of the modern river Isla upstream from Cardean. Those who used the inversion-type hill-letter r also applied their river-letter s in an inversion-type manner to the minor stream, i.e. they placed their river-letter s before the river-letter l, thus creating the river-name which became the modern Isla.
5 But there seems to have been a changeover period in which although hill-letters were used in an inversion-type manner, the corresponding river-letters continued to be used in the old-fashioned manner, i.e. in chronological order going from left to right within the river-name. For example, the main river at Smetriadum (Bainbridge) was the Isur (now the Ure, which flows close by the Bainbridge fort). The chronological order of the hill-letters is first r, then m and finally s. The earliest hill-letter is thus r, corresponding to the river-letter s. The river-letter r corresponding to the inversion-type hill-letter m is, however, placed after the s in the river-name Isur, i.e. the river-letter r was still applied in an old-style manner even though the corresponding hill-letter m is used in an inversion-type manner. The river-letter b, corresponding to the hill-letter s, is present in the name of the river Bain, which flows past the fort and joins the Ure just north of Bainbridge. (But see the entry for Smetriadum in the Alphabetical List, where it is suggested that the Celtic place-name might have been Smedriadunum and the river Ure might have been the Urus). In addition, reference is made to the Ravenna place-name Ugueste. By comparing the structure of Ugueste with that of Ugrulentum one can see that two hill-letters are missing from Ugueste. Now, Ugueste at Stirling stood on the river called Bdora by Ravenna and Boderia by Ptolemy. The river-letters are thus b, t (changed to d) and r and the corresponding hill-letters are s, l and m. Since s is the last hill-letter in the place-name and b the first river-letter in Bdora, it follows that s must actually be the earliest hill-letter in the place-name and b of course the earliest river-letter in the river-name. The river-letters are thus still arranged in the old-style manner, i.e. in the chronological order first b, then t, changed to d, and finally r, even though the elements of Ugueste before the st element must be inversion-type. Another example is afforded by the place-name Vresmedenaci (Celtic form of Bresnetenaci/Bremetonnaci) at Ribchester and the river Berisama (Celtic form of Ptolemy's Belisama), now the Ribble. The hill-letter r of Vresmedenaci is used in an inversion-type manner, but in the river-name Berisama the corresponding river-letter s appears after the river-letters b and r corresponding respectively to the old-style hill-letters s and m of Vresmedenaci.
6 There is a further complication in that those people who used the hill-letter l2 appear normally to have placed their river-letter t at the end of an existing river-name. Thus the chronological order of the hill-letters in Lecimocsava (Ravenna’s Levioxava) is first s, then m and finally l, the m and l both being used in an inversion-type manner. But in the river-name Nethy the river-letter t corresponding to the hill-letter l is placed after the river-letter b corresponding to the hill-letter s. Note that in this case the b changed to v and then the v switched to n. The river-name corresponding to the place-name Lecimocsava will thus have had a form such as Arbet or Arbete, which, with b→v→n, became Arnet or Arnete. Presumably later settlers coined the name Aberarnete for their settlement at the mouth of the river Arnete and Aberarnete was later shortened to Abernete, which later became the modern Abernethy. The nethy element of Abernethy thus has no connection with any "Pictish" king called Nectan or Nechtan, as is commonly supposed. Note that it was the initial part Arne of the river-name which survived as the modern river-name Earn. Furthermore, the Celtic name of the fort at Cardean was Rugulentum, the chronological order of the hill-letters being first n, then l, and finally r. The earliest hill-letter in the place-name is thus n, corresponding to the river-letter m applied to major rivers. The people who used the hill-letter l, in fact l2, then placed their river-letter t after the m, thus creating the river-name Matovion, which was apparently applied to that stretch of the modern Isla between Cardean and the confluence with the Tay (the river-name was then transferred by the Romans to their fort on the banks of the river at Cargill, Matovion being the Ravenna name immediately before Ugrulentum). One sees this river-letter t at the end of a number of river-names, for example in Derbentione and Trisantonis and in the modern river-names Dart, Teviot, Tweed (t changed to d) and Teith (t changed to th). The river-letter t at the beginning of all of these names (changed to d in Derbentione and Dart) corresponds to the hill-letter l1.
7 With only five exceptions all of the river-names in Ptolemy and Ravenna can be assigned to four categories. These are:
7.1 River-name comprising only one or more river-letters (leaving aside name-endings)
Abi, Abona (242), Adron (254), Alaunus/Alavna (263), Anava (251), Aventio (245), Bdora (252)/Boderia, Berisama, Deva, Devas, Devion, Devoni, Dorvantium (250), Duabs, Durbis (265), Eltabo, Iena, Ila, Itis, Ituna, Liar (259), Matovion, Novitia (253), Novius, Rumabo, Rutupis, Sabrina, Seteia, Tamaris (240)/Tamarus, Tamesa, Tamion (244), Tina, Tinea (258), Tobius, Toesobis, Trisantonis, Tuerobis, Vedra, Veromo, Vividin (261), Volas bay
7.2 Topographical land-name of a place located on or close to a river or bay simply transferred by the Romans to the river or bay
Certisnassa (255), Gabrantuicorum bay, Isca (243), Isca(s), (Is)caelis, Lemana (266), Lemannonius bay, Lenda (260), Longus, Loxa, Metaris, Sidumanis, Stuccia, Tuessis/Tuesis, Uxella
Dunum bay probably belongs here. This is based on the duno element of Lugunduno (140), though it is not clear whether only this element was transferred to the river, now the Tees, or whether originally the entire name was so transferred.
7.3 River-element, comprising one or more river-letters, added as a prefix to a land-name
Abra vannus, Duro lavi (262), Mori cambe, Na barus (from Ba barus), Re rigonius, Tr axula (236), Ve lox (271)(from Be locs)
7.4 River-element, comprising one or more river-letters, added as a suffix to a land-name
Anderelion uba, An trum (257), Arduara venatone, Can ubio, Coan tia (249), Coguve usuron (264), Gar iennus, In traum (256), Leugo sena (248), Mavia sarna (238, 239), Rato stabius, Raxtomessa senna (268, 269), Var ar (from Bar ar)
The Clota may belong here. It appears to comprise the cel of Cindocelum (204)(assuming there should be only one l in this name) with the river-letter t as a suffix. What is not clear is whether the river-letter t was added only to the cel element or whether it was originally added to the whole place-name. However if, as seems possible, the Celtic place-name was originally Cindoceldum, then Ptolemy's Clota may just be a modification of the celd element of the place-name.
8 River-names and bay-names not included in the categories set out above:
Ptolemy: Cenio
Ravenna: Naurum (241), Bribra, Novia (267), Cunia (270)
Cenio and Cunia may be category 2 names, if the n is the hill-letter n, both names then meaning ‘steep hill’. Bribra is a category 3 or 4 name, though it is not clear which. Naurum is probably a category 1 name, but this is not sufficiently clear. Novia appears to be a modified form of the nuba of Anderelionuba.
9 And now an attempt is made to identify all of the river-names given by Ptolemy and Ravenna for the whole of Britain. Table 1 gives in the left-hand column the river-names provided by Ptolemy, in the right-hand column those provided by Ravenna, and in the middle column the corresponding modern river-names. The names are arranged in geographical order going clockwise round Britain starting with the river Traxula, the Axe in Devon, as Ravenna does. Names in brackets are names which do not actually appear in the respective river list - they appear in the place-name list but are river-names which were adopted by the Romans and applied by them to forts built on the banks of the rivers concerned. Where a particular name in either the Ptolemy or Ravenna column cannot be geographically ordered with respect to the names in the other column it is marked with an asterix. The asterix merely alerts the reader to the fact that a name so marked can only be located relative to the immediately preceding and succeeding names in the same column (ignoring names in brackets). The various promontories referred to by Ptolemy are included in Table 1, not because identifying them is of any relevance to the present study, but because they sometimes help to ensure that a particular river or group of rivers is placed in the right region of the country.
Table 1
|
Ptolemy |
Modern name of river or bay |
Ravenna |
|
|
|
Axe |
Traxula (236) |
|
|
Exe |
Axium (237) |
||
|
|
Unknown, but see note |
Mavia (238) |
|
|
|
Unknown, but see note |
Sarna (239) |
|
|
Tamarus |
Tamar |
Tamaris (240) |
|
|
Cenio |
Unknown |
|
|
|
Damnonium promontory Ocrium promontory |
Presumably Lizard Point |
|
|
|
Antivestaeum promontory Bolerium promontory |
Presumably Land‘s End |
|
|
|
|
Camel |
(Eltabo) |
|
|
Neet |
(Purocoronavis) |
||
|
Okement+Torridge(part) |
|||
|
|
Taw |
(Devion) |
|
|
Herculis promontory |
Presumably Hartland Point |
|
|
|
Parrett estuary |
|
||
|
Sabrina estuary |
Severn estuary |
|
|
|
|
Unknown, but see note |
*Naurum (241) |
|
|
|
Avon (Bristol) |
Abona (242) |
|
|
|
Avon (Stratford) |
(Alavna) |
|
|
Usk |
Isca (243) |
||
|
(Ratostabius?) |
Taff |
Tamion (244) |
|
|
|
Ewenny |
Aventio (245) |
|
|
|
(Error when copying Leugosena) |
Leuca (246) |
|
|
|
(See note) |
Iuctius (247) |
|
|
|
Loughor |
Leugosena (248) |
|
|
|
|
|
|
|
Towy |
|
||
|
Octapitarum promontory |
Promontory on Pembroke coast |
|
|
|
Teifi |
|
||
|
Ystwyth |
|
||
|
Dovey |
|
||
|
Caeanganorum promontory |
Promontory on Lleyn peninsula |
|
|
|
Seteia estuary |
Mersey estuary |
|
|
|
Belisama estuary |
Ribble estuary |
|
|
|
Setantiorum harbour |
at the mouth of the river Wyre |
|
|
|
Moricambe estuary |
Lune estuary |
|
|
|
Kent |
(Cantiventi) |
||
|
|
Duddon |
Coantia (249) |
|
|
|
Esk |
(Alavna) |
|
|
|
Lowca Beck |
(Bribra) |
|
|
|
Derwent |
Dorvantium (250) |
|
|
Ituna estuary |
Solway Firth |
|
|
|
|
Annan |
Anava (251) |
|
|
Nith |
|
||
|
Devas |
Dee |
|
|
|
Iena estuary |
Estuary of Water of Fleet |
|
|
|
Water of Luce |
|
||
|
Novantarum promontory |
|
|
|
|
Novantarum peninsula |
|
|
|
|
Rerigonius bay |
Girvan bay |
|
|
|
Vindogara bay |
Irvine bay |
|
|
|
Clota estuary |
Firth of Clyde |
|
|
|
Lemannonius bay |
Lamlash bay, Arran |
|
|
|
Epidium promontory |
|
|
|
|
Add |
|
||
|
Unknown, but see note |
|
||
|
Volas bay |
Apparently the bay at the entrance to Loch Eriboll in Sutherland |
|
|
|
Nabarus |
Naver |
|
|
|
Tarvedum promontory Orcas promontory |
Holborn Head (Caithness) |
|
|
|
Virvedrum promontory |
St John's Point (Caithness) |
|
|
|
Verubium promontory |
Duncansby Head |
|
|
|
Ila |
Helmsdale |
|
|
|
Ripa alta |
|
|
|
|
Varar estuary |
Beauly Firth |
|
|
|
Lossie |
|
||
|
Taezalon promontory |
|
|
|
|
Deva |
Dee |
|
|
|
|
Bervie Water |
(Iberban) |
|
|
South Esk |
|
||
|
Tuesis estuary |
Firth of Tay |
Tay |
(Tuessis) |
|
|
Isla, between Cardean and the Tay |
(Matovion) |
|
|
|
Almond (in Perthshire) |
(Veromo) |
|
|
Tina |
Eden (in Fife) |
|
|
|
|
Devon |
(Devoni) |
|
|
Boderia estuary |
Firth of Forth |
Forth |
Bdora |
|
|
Almond (in Lothian) |
(Rumabo) |
|
|
|
Tyne (in East Lothian) |
Novitia (253) |
|
|
|
Whiteadder |
Adron (254) |
|
|
|
Tweed |
Certisnassa (255) |
|
|
|
(Duabs) |
||
|
|
Unknown, but see note |
Intraum (256) |
|
|
|
Unknown, but see note |
Antrum (257) |
|
|
|
Eden Water |
Tinea (258) |
|
|
|
Yarrow |
Liar (259) |
|
|
|
Till |
Lenda (260) |
|
|
|
Teviot |
Vividin (261) |
|
|
|
Lyne Water |
Durolavi (262) |
|
|
Alaunus |
Aln |
Alavna (263) |
|
|
|
Coquet |
Coguveusuron (264) |
|
|
Wear |
|
||
|
Dunum bay |
Tees bay |
|
|
|
Gabrantuicorum bay with many harbours |
Humber |
|
|
|
Ocelus promontory |
|
|
|
|
Witham |
|
||
|
Metaris estuary |
Estuary of Great Ouse |
|
|
|
Yare |
|
||
|
Promontory? |
|
|
|
|
Blyth |
|
||
|
Tamesa estuary |
Thames estuary |
|
|
|
Cantium promontory |
|
|
|
|
(Rutupie) |
Stour |
(Rutupis) |
|
|
|
Dour |
Durbis (265) |
|
|
|
East Rother |
Lemana (266) |
|
|
|
River at Pevensey |
Novia (267) |
|
|
Novus Portus (New Harbour) |
Unknown |
|
|
|
|
Adur (formerly called Sore) or Ouse |
Raxtomessasenna (268, 269) |
|
|
Arun, formerly called Tarrant |
|
||
|
|
Unknown, but see note |
*Cunia (270) |
|
|
Magnus Portus (Great Harbour) |
Portsmouth Harbour or Southampton Water |
|
|
|
Alaunus |
Avon (in Hampshire) |
||
|
Brit |
Velox (271) |
||
10 It might help the reader understand British Celtic river-names and their modern descendants more clearly if it is emphasised here that it is simply misguided to invent hypothetical early Celtic words to explain the meanings of those river-names. British Celtic river-names had no meaning. Each name simply developed by a process of accretion as each of successive waves of settlers added its own river-letter to the already existing river-name. The final form of such a river-name thus depends on a number of factors:
- which particular groups of settlers actually lived along the course of the river;
- the chronological order in which the groups of settlers arrived;
- whether a group of settlers added its river-letter to the existing river-name before about 130BC (when old-style place-names were being coined) or after about 120BC (when inversion-type place-names were being coined; and
- the changes which particular consonants experienced: thus the river-letter b might change to a v (as in the modern names Severn, Dove and Tove), to an n (as in Earn and Nith), to a w (as in Taw) or even an f (as in Fleet): and the river-letter t might change to a d (as in the modern names Eden and Don), to th (as in Thames, Witham, Teith and Neath) or even dd (as in Nidd, Duddon and Add).
Thus, for example, if we imagine an m-people (a people who used the hill-letter m) applying their river-letter r to a river, and then, sometime after 120BC, an l1-people added their river-letter t to the front of the existing river-name, we obtain the common river-compound tr seen in Traxula and Trisantonis and, with the t changed to d, in Adron and Derventio. An l2 -people might then arrive and add their river-letter t to the end of the existing river-name. They might add it only to the proper-name part of the river-name (as in Dart) or to the earlier river-name complete with its on/en/in ending (as in Darent, Trisantonis and Derventio). If we start further back in time and imagine an s -people applying their river-letter b to a river and then, after 120BC, an l1-people added their river-letter t to the front of the existing river-name, we obtain a modern river-name such as Deben (where the t has changed to d) and also Dove and Tove, where the b has changed to v. If an m-people had settled somewhere along the river after the s -people, later than 120BC but before the l1 -people, we would obtain the river compound Terben, which yields the modern river-name Darwen, where the t has changed to d and the b to v, the latter then changing to w. If an l2 -people added their river-letter t to the end of the whole element Terben, then with the t→d change we obtain the fairly common river-name Derbent, sometimes appearing with an io ending, sometimes an ione. With the b changed to v we have the better-known form Derventio. The name Trisantonis developed along the same lines, though here the starting point is the river-letter r applied by an m-people, then came the river-letter s applied, before 130BC, by an r-people, then initial river-letter T applied, after 120BC, by an l1-people and finally the second river-letter t applied by an l2 -people. Finally, the river-name Sabrina started out with an s-people applying their river-letter b, then an m-people (actually the predecessors of the Ordovices in that region: we see their hill-letter m in the Med element of Mediolano at Whitchurch, the l in Mediolano being added later by the Ordovices themselves) applied their river-letter r, and finally, at some point later than 120BC, an r-people (either the Cornovi or the Dobunni) added their river-letter s to the front of the existing river-name, the river-name now becoming Sabrina. One could consider other examples, but the process of creating a Celtic river-name and then developing it by adding new river-letters should by now be quite clear.
11 The river-names listed in paragraphs 7.3 and 7.4 above repay further study since although some of those names may appear to be haphazard collections of letters there was in fact method in the creation of those names. Those river-names, each a compound of a river-element and a place-name element, are here referred to as composite river-names. A river-element appearing at the front of a composite river-name is here referred to as a river-prefix and a river-element appearing at the end of a composite river-name is here referred to as a river-suffix. But it should be noted that in all of those names it is in fact the river-element which is the generic element in the composite river-name and it is qualified by the place-name element. The names serve to distinguish between two rivers of the same name. Thus, to take two examples from the list in 11.1 below, there was a river Ventena in Cumbria (the Kent) and another in Northamptonshire (the Nene). In order to distinguish these rivers clearly from one another the river in Cumbria was called the Ventena which flows past the place called Candion, the composite river-name then being written down as Candiventena, modified by the Romans to Cantiventi. And the river in Northamptonshire was called the Ventena which flows past the place called Bandion, the composite river-name being written down as Bandaventena, modified by the Romans to Bannaventa. In order to have more material to study the names listed in paragraphs 7.3 and 7.4 above are supplemented here by river-names which served as a basis for tribal names (listed in Chapter 27, 4) and one or two place-names which are clearly based on river-names. Elsewhere on this website the term ‘old-style’ is applied to topographical place-name elements in which a qualifying letter, such as b meaning ‘high’ or c/g meaning ‘steep’, comes before the hill-letter in the element. The term ‘inversion-type’ is applied to topographical place-name elements in which the qualifier, in this case t meaning ‘high’ or c/g meaning ‘steep’, comes after the hill-letter. In the present section the use of the term ‘old-style’ is extended to cover a river-letter which is added to the end of an existing river-name, or is placed after a qualifying place-name in a composite river-name, and the term ‘inversion-type’ is applied to a river-letter which is added to the front of an existing river-name or is placed before a qualifying place-name element in a composite river-name. Likewise an old-style composite river-name is one where the qualifying place-name element comes before the generic river element in the river-name and an inversion-type composite river-name is one in which the qualifying place-name element comes after the generic river element in the river-name. One last point to bear in mind is that the normal chronological order of the hill-letters in old-style place-names is n1, s, m, r, l1, n2. The normal chronological order of the corresponding river-letters in old-style river-names is l/m, b, r, s, t, l/m. Clearly one has to be extra careful with the hill-letter n in place-names and with the corresponding river-letters l and m in river-names. The hill-letter l in an old-style place-name element will be l1 since the people who used the hill-letter l2 appear only to have coined inversion-type place-names. In an inversion-type place-name it is not always clear whether the hill-letter l is l1 or l2, but in a corresponding river-name the l1-people seem always to have placed their river-letter t at the front of an existing river-name or element, whereas the l2 -people seem always to have placed their river-letter t at the end of an existing river-name or element.
11.1 The earliest composite river-names are old-style names comprising an old-style place-name qualifying an old-style generic river-element. Examples include:
|
Arduaravenatone (Arduaravenatone) (river Okement) |
Place-name old-style Ard with ara ending. There may have been a hill-letter l between the u and the a. River-suffix venat includes river-letters b (changed to v) and t corresponding respectively to hill-letters s and l. |
|
Bandaventena (Bannaventa) (river Nene) |
Place-name old-style Band in hill-letter n1. River-suffix vent includes river-letters b (changed to v) and t corresponding respectively to hill-letters s and l. |
|
Barisena (Parisi) (river Seine) |
Place-name old-style Bar in hill-letter r. River-suffix sena includes river-letter s corresponding to hill-letter r. |
|
Bindomi (Vindomi) (river Wey) |
Place-name Bind in hill-letter n1. River-suffix comprises river-letter m corresponding to hill-letter n1. |
|
Bindomora (Vindomora) (river Derwent, County Durham) |
Place-name Bind in hill-letter n1. River-suffix mora includes river-letters m and r corresponding respectively to hill-letters n1 and m. Alternatively place-name Bindom in hill-letters n1 and m. River-suffix comprises river-letter r corresponding to hill-letter m. |
|
Candiventena (Cantiventi) (river Kent, Cumbria) |
Place-name Cand in hill-letter n1. River-suffix vent includes river-letters b (changed to v) and t corresponding respectively to hill-letters s and l. |
|
Carventena (Carveti) (river Eden, Cumbria) |
Place-name Car in hill-letter r. River-suffix vent includes river-letters b (changed to v) and t corresponding respectively to hill-letters s and l. |
There are composite river-names similar to those listed above but where an inversion-type element in the hill-letter l2 was added to the place-name. The corresponding river-letter t was simply added to the end of the generic river-element, just as l2 -people normally added their river-letter t to the end of an existing river-name. Corionototena is an example of such names (see Chapter 27 for a discussion of this name). Lutecstobertena/Lutecsdobertena is another example (this name is also discussed in detail in Chapter 27). Then there are two composite river-names in which the river-letter b is used in the same manner as the river-letter t referred to above (some people who used the hill-letter s appear to have been somewhat dilatory in changing over to inversion-type river-names, even though they had changed over to inversion-type place-names). The two names are Canubena and Anderelionuba. Canubena (upon which is based the place-name Canubio of the Roman fort at Caerhun, this point being discussed in more detail in the notes for the Deceangli tribe in Chapter 27) appears to be a simplified version of Descecanglubena, apparently the Celtic name of the river Conwy in North Wales. The second name, Anderelionuba, also has as river-element the river-letter b corresponding to the hill-letter s, this suggesting that the place-name had in fact been Sanderelion, where the hill-letter s is inversion-type. But here again the river-element may earlier have included river-letters corresponding to the hill-letters n1, r and l in Anderelion. Letucarosena (Leugosena) is a name of the same general kind as those listed above, the river-suffix sena including the river-letter s corresponding to the hill-letter r in the place-name element. In this case, however, it is not clear whether the hill-letter l in the place-name is l1 or l2, but in either case the corresponding river-letter t has not been added to the river-suffix sena.
11.2 Following the changeover from old-style to inversion-type names, in some composite river-names the river-element was simply moved from the end of the river-name to the front, without altering the order of the river-letters, so that the new composite river-name included the generic river-element followed by the qualifying place-name element. Examples include:
|
Abranobandion (Abravannus) (Water of Luce, Dumfries and Galloway) |
Place-name Band in hill-letter n1. River-element (originally a suffix) includes river-letters b and r corresponding to hill-letters s and m. Not clear if there had been a river-letter l/m (corresponding to hill-letter n1) at the front of the river-element. The Novantae used the river-letter r, so it will be they who moved the river-element to the front of the river-name. |
|
Burocoronavis (Purocoronavis) (river Neet, Cornwall) |
Place-name Coronavis in hill-letters r and n2. River-element bur includes river-letters b and r corresponding to hill-letters s and m. |
|
Eburocaslum (Eburocaslum) (river Esk, Dumfries and Galloway) |
Place-name Caslum in hill-letters s and l. River-element ebur includes river-letters b and r corresponding to hill-letters s and m. |
|
Ortobicsena (Ordovices) |
Place-name bics in hill-letter s. River-element (again originally a suffix) includes river-letters r and t corresponding to hill-letters m and l. The Ordovices used the hill-letter l (actually l1), so it will be they who moved the river-element to the front of the composite river-name. |
|
Velocs (Velox) (river Brit, Dorset) |
Place-name Locs in hill-letters s and l2. Original river-suffix included river-letter b (corresponding to the hill-letter s) and may have included river-letters r and t (as in modern river-name Brit). River-element moved to front, b changed to v, river-letters r and t lost or dropped. |
11.3 In other cases the river-element remained at the end of the composite river-name but was itself made inversion-type by having one or more river-letters placed in inversion-type manner at the front of the element. Examples include:
|
Antrum (Leader or Ettrick, Scottish Borders) |
Place-name lost or omitted apart from ending an. River-letter r earlier river-letter in river-element, the river-letter t being inversion-type (applied by Votadini). |
|
Garisernus (Gariennus) (river Yare, Norfolk) |
Place-name Gari with river-suffix earlier of form resnus. River-letters r and s correspond to hill-letters m and r which appear in this chronological order in that region, as in Combredovio and Camborico. River-letter s moved to front of river-suffix by the Iceni. |
|
Intraum (Ettrick or Leader, Scottish Borders) |
Same as Antrum. Place-name ending in this case in. Not clear if there had been another river-letter between the a and the u. |
|
Lacerocomaguvetusuron (Coguveusuron) (river Coquet, Northumberland) |
Inversion-type river-suffix tusuron corresponding exactly with place-name Lacerocomaguve. Earliest hill-letter m, earliest river-letter r (some people who used the hill-letter m, present here in the mag element, were rather dilatory in changing over to inversion-type river-names). |
|
Racstomessaserna (Raxtomessasenna) (river Ouse, East Sussex) |
Place-name Racstomessa plus river-suffix senna. Hill-letter m in place-name. Corresponding river-letter r probably in river-suffix. Hill-letter r inversion-type. Corresponding river-letter s placed at front of river-suffix serna. |
|
Ratostabius (river Usk or Taff?) |
Inversion-type place-name Rat and river-suffix stabius. Inversion-type river-letter s corresponds to hill-letter r. Place-name element may earlier have included hill-letters l and s corresponding to river-letters t and b. |
Maviasarna will be a name of the same kind as those listed above. The river-element sarna includes the river-letter r corresponding to the hill-letter m in the place-name Mavia. The river-element also includes the inversion-type river-letter s but the corresponding inversion-type place-name element in the hill-letter r is missing from the name Mavia. Of course it is possible that people who used the hill-letter r lived somewhere else along the course of the river, but not at Mavia itself. The river-element could then change, but the place-name Mavia would not.
11.4 Some composite river-names were a further development of the kind discussed in 11.3 above in that the modified river-element, itself now inversion-type, was moved to the front of the composite river-name so that this name as a whole was now inversion-type, i.e. generic river-element followed by qualifying place-name. Examples include:
|
Atrebandena (Atrebates) |
Old-style place-name Band in hill-letter n1. River-element includes river-letter r corresponding to hill-letter m. River-letter t is inversion-type. There was probably an l/m at front of river-element corresponding to hill-letter n2 since the Atrebates used the hill-letter n2. |
|
Durolavi (Lyne Water, Scottish Borders) |
Earlier river-letter in river-element Dur is r. The river-letter t (changed to d) is inversion-type. River-element moved to front by Votadini, who used the hill-letter l1 corresponding to river-letter t. The lavi element may be a shortened version of a longer place-name including the hill-letters l and m. |
|
Moricambe (river Lune) |
Old-style place-name Cambe in hill-letter m (part be of place-name is incomplete or unclear). Corresponding river-letter r in original river-suffix. River-letter m (corresponding to hill-letter n2) of river-suffix is inversion-type. River-suffix moved to front of composite river-name. River-letter m Indicates that n2 -people (presumably the Brigantes) took over part of river-valley, probably around the confluence of the Leck Beck with the Lune. |
|
Tabecsali (Taexali) (river Bogie, Aberdeenshire) |
May be river-suffix t moved to front of place-name Becsali. Alternatively may have been place-name Ecsali plus river-suffix bat comprising river-letters b and t corresponding to hill-letters s and l of Escali. River-letter t shifted to front of river-element and river-element moved to front of place-name Escali, thus yielding composite river-name Tabecsali. This river-name is the basis for the tribal name Taexali. |
|
Tracsula (Traxula) (river Axe, Devon) |
River-letters r and t correspond to hill-letters m and l. People who used river-letter t and hill-letter l (presumably Durotriges) moved the river-letter t to front of river-element and then moved entire river-element to front of composite river-name. |
|
Trinobandena (Trinobantes) (river Pant/Blackwater, Essex) |
Place-name Band in hill-letter n1. Original river-suffix rt with river-letters r and t corresponding to old-style hill-letters m and l, seen in Camul (Trinovantian name for Colchester). River-letter t moved to front of river-element and entire river-element moved to front of composite river-name. River-name is origin of tribal name Trinobantes/Trinovantes. |
11.5 There were of course composite river-names which appear to have been inversion-type from the beginning. Only two examples come to mind:
|
Babarus (Nabarus) (river Naver, Sutherland) |
River-letter b and old-style place-name Bar in hill-letter r. River-letter b applied by the Smertae, who coined inversion-type place-names in the hill-letter s. At some stage the initial b changed to n. |
|
Rerigonius (Rerigonius) (Water of Girvan, Ayrshire) |
Generic river-letter r qualified by inversion-type place-name Rigonius in hill-letter r. River-letter applied by the Novantae. |
11.6 It may be noted, in conclusion, that the composite river-name Barar (Ptolemy’s Varar estuary, the Beauly Firth) is an example of names showing that some people who used the hill-letter m and river-letter r were rather slow to change over to inversion-type river-names. The Lugi, who used the hill-letter m, appear to have moved to Easter Ross after the changeover to inversion-type place-names (see ‘The Celtic “Picts”: Part 1’, 5.7) and yet they continued to use their river-letter r in the old-style manner, i.e. they placed it after the place-name element Bar in Barar.
[This page was last modified on 08 October 2024]
