[NB. It will probably be easier for the reader to understand this page if he or she is already familiar with the contents of at least Chapters 1 and 2. Chapter 1 deals with the basic building-blocks of Romano-British place-names. Chapter 2 explains how the basic building-blocks were assembled by the Celts to form compound names.]



Chapter 24

Summary and conclusions


1             We have covered a lot of ground on this website. What began as a study of Romano-British place- and river-names ended up taking us much further back into Britain’s history, for it is clear that many of the place-names in Ravenna and Ptolemy were originally the names of Celtic settlements which existed long before the Romans set foot in Britain.

2             We have seen that several groups of settlers arrived in Britain from the Continent and that they used different letters of the alphabet to signify ‘hill’ and also different letters to signify ‘river’. In the interests of simplicity all of those settlers are here regarded as Celts.  The following table shows the correspondence between the hill- and river-letters used by the different groups of settlers.

Hill-letter Corresponding river-letter





n m Lindum Witham
    Londinium Tamesis (Thames)
    Magno Frome
  l Camulodono Eller Beck
    Lindum Till
    Rugulentum Isla
s b Lucamosessa Derbentio (Derwent)
    Mugulesde Bdora/Boderia (Forth)
    Ucsella Parrett (b→p)
m r Elconionemedo Tamaris (Tamar)
    Lucamosessa Derbentio (Derwent)
    Magno Frome
r s Averno Stour
    Eburacum Ouse
    Vertis Sabrina (Severn)
l t Elconionemedo Tamaris (Tamar)
    Lindum Till, Witham (t→th)
    Londinium Tamesis (Thames)


(Note that the river-letter m appears to have been used for a main river, whereas the river-letter l was used for a tributary or headwater of a main river. For example, the upper reaches of the modern river Isla, upstream from Cardean, evidently had a name substantially the same as Isla, including the river-letter l, whereas the main river between Cardean and the confluence with the Tay was clearly called Matovion, this including the river-letter m). There is a slight complication in that the hill-letters n and l were each used by two different groups of settlers who arrived in Britain at different times.  There were thus actually seven different groups of settlers and the normal chronological order of the hill-letters, and thus the chronological order in which the settlers arrived in Britain, was n1, s, m, r, l1, n2, l2. There is, however, some regional variation, as in Lincolnshire, where the hill-letter n2 arrived before the hill-letter l1, as seen in the old-style place-names Croconcalana (later Crococalana) and Banvobalum (later Bannovalum). The settlers who used the hill-letters n1 to n2 were all in Britain by 150BC, as evidenced by the history of Maiden Castle in Dorset, explained in Chapter 23, 3.3. The l2 people appear to have used only inversion-type names and so arrived in Britain later.

3             We have seen that Celtic place-names fall into three broad categories, namely old-style names, transitional names and inversion-type names. In old-style place-names any qualifier comes before the hill-letter in the name or the hill-letter is itself used as a qualifier for the letter d, signifying ‘summit’, or v, signifying ‘slope, side of a hill’, the hill-letter coming before the d or v in the name. In transitional names the hill-letter has two qualifiers, one coming before the hill-letter, the other after. In inversion-type names the qualifier comes after the hill-letter or the hill-letter is itself used as a qualifier for the letter d, again meaning ‘summit’, or the letter v, again meaning ‘slope, side of a hill’, the hill-letter coming after the letter d or v in the name.

3.1          In some old-style names the letter b, meaning ‘high’, qualifies the hill-letter, as in Banva (later Banna), Banvobalum (later Bannovalum), Bereda and Binovia (later Vinovia). In other names the letter c (sometimes replaced by g), meaning ‘steep’, is used in the same way, as in Caluvio, Cerma, Clindum and Credigone. And in some names both b and c are used, as in Becsa (later Pexa), Bograndium, Cibra and the bagl element of Cambaglanda (later Gabaglanda, later still Amboglanna). Where an old-style name includes the letter d meaning ‘summit’ the hill-letter comes before the d, as in the and element of Manduesedo or Anderelion (a place-name included in the river-name Anderelionuba). In some old-style names of this kind the hill-letter may itself by qualified by b and/or a c as discussed above, as in Bereda, Bindogladia, Condecor and Corda. Where an old-style name includes the letter v meaning ‘slope, side of a hill’, the v is used in the same way as the d, meaning ‘summit’, discussed above, examples being Alvinundo (later Albinumno), Bamvocalia (later Pampocalia), Banva (later Banna) and Calveva (later Calleva). 

3.2          There are only a few transitional names and in these names the hill-letter is qualified by both ‘high’ and ‘steep’. Where the qualifier ‘high’ comes before the hill-letter it is represented by b, as in Biriconion (later Viriconium) and Brocara. Where it comes after the hill-letter it is represented by t, as in Cantia (later Canza) and Cunetio. The qualifier ‘steep’ is represented by c (or g) no matter whether it comes before the hill-letter, as in Cunetio, or after, as in Brocara.  

3.3          In inversion-type place-names the two qualifiers ‘high’ and ‘steep’ come after the hill-letter in the name, though ‘high’ is represented by the letter t rather than by b as in old-style names. Thus in an inversion-type name ‘high hill’ is written as ‘hill high’, as in Litana, Vertis, the ent element of Lagentium and the lut element of Lutudaron. ‘Hill steep’ is represented by a hill-letter followed by a c (or g) as in Alicuna, Isca and the lag element of Lagentium and Lagubalium. In some names both c and t are present, as in Anicetis and Macatonion. ‘Hill summit’ is represented by a d followed by a hill-letter, as in the dun element of Camulodunum, the dar element of Lutudaron or the Dem element of Demerosesa.  If the hill is high the letter sequence is d, then a hill-letter and then a t, as in the dert element of Omirededertis (later Omiretedertis). If the hill is steep the letter sequence is d, then a hill-letter and then a c (or g), as in the denac element of Vresmedenaci (later Bresnetenaci). A location on the side of a hill is represented by a v followed by a hill-letter, as in the ven element of Venta, the ver element of Verulamium and the vel element of Velurcion.

4             The changeover from old-style to inversion-type place-names appears to have occurred during the second half of the second century BC, as explained in Chapter 23, 6.1. The change will not have occurred overnight and we may assume that the transitional names were coined during the period of the changeover. In the interests of simplicity and brevity we will assume that old-style names were coined prior to 130BC, that transitional names were coined in the ten-year period from 130 to 120 BC and that inversion-type names were coined after 120 BC. The l2-people appear to have used only inversion-type names and so apparently arrived after 120BC. But they may have arrived a little earlier, since it may have been the very arrival of these people which triggered the changeover from old-style to inversion-type names on the part of the other tribes already living in Britain.

5             There are about 240 place-names in Ravenna and Ptolemy, where two names of the same form but clearly relating to different places are counted as two place-names, and two names of different form in Ravenna and Ptolemy but clearly relating to one and the same place are counted as one place-name.  The eight diversa loca of Ravenna are included for completeness.  About 71% of the place-names are topographical names referring to a hill, a high hill, a steep hill, the summit of a hill or a location on the side, the slope of a hill.  About 15% of the place-names are in fact river-names transferred to a Roman fort (and its vicus) or a Romano-British settlement (Chapter 2, 4.1.15). Another 5% of the place-names are Latin names or names with Latin elements (Chapter 2, 4.2.1). About 58% of the topographical names are old-style names or names with old-style elements (these are the names indicated in Chapter 2, 4.1.1 – 4.1.4, 4.1.6, 4.1.10, 4.1.11, 4.1.13, 4.1.14). Some 44% of the topographical names are inversion-type names or names with inversion-type elements (Chapter 2, 4.1.7 - 4.1.14). Only about 6% of the topographical names are transitional names or names with transitional elements (Chapter 2, 4.1.5, 4.1.6, 4.1.8).

6             As noted above some 58% of the topographical names are old-style names or names with an old-style element (compounds with a transitional or inversion-type element) and these old-style names and old-style elements were all in use prior to 130BC. And in the case of an old-style compound name, comprising two or more hill-letters, the earlier or earliest element of the compound may have been coined significantly earlier than 130BC. The fact that the old-style names and old-style name elements survived long enough to be included in Ravenna and Ptolemy suggests continuity of occupation through the late Iron Age and into the Roman period, but this does not necessarily mean that the earlier Celtic settlement was on exactly the same site as the later Roman fort (and its vicus) or Romano-British settlement. In many cases the Romans transferred the name of a hillfort, and presumably its inhabitants, to a new settlement on a more convenient site nearby. For example the name and presumably the inhabitants of the hillfort in Credenhill Park Wood in Herefordshire were transferred to the new Magnis at Kenchester. Likewise, the name and presumably the inhabitants of the hillfort on Crickley Hill in Gloucestershire were transferred to the new Glevon at Kingsholm in the northern outskirts of Gloucester, and the name and presumably the inhabitants of the hillfort on Barcombe Hill in Northumberland were transferred to the new Bindolande (later Vindolande) at Chesterholm.

7             One finds place-names in all hill-letters in most parts of Britain, but by the date of the Roman invasion in AD43 particular hill-letters dominated in particular parts of the country, this indicating that particular groups held sway in particular parts of the country. It has been possible to associate the tribal names given by Ptolemy with some of those groups. No area of Britain has been identified in which a group using the hill-letter n1 held sway at the date of the Roman invasion, but a group using the hill-letter s appears to have dominated in north Wales. These people were presumably the tribe generally known as the Deceangli. Ptolemy’s Dumnoni in southwest England and Damnoni in central Scotland used the hill-letter m. The Iceni, the Dobunni and the Cornovii  used the hill-letter r. The Trinovantes, the Coritani (Corieltauvi), the Durotriges and the Silures all used the hill-letter l1. The Atrebates, the Belgae, the Regni, the Catuvellauni and the Brigantes all used the hill-letter n2. The Parisi  appear to have used the hill-letter l2. These points are discussed in detail in “Ptolemy’s Celtic tribes in Britain”, accessible from the main menu above.

8             It has also been possible to demonstrate, using the place-name evidence, that there was tribal movement within Britain. This has only been done for southeast England and the north of Scotland, but it should be possible to carry out similar studies for other parts of Britain. It has thus been possible to show that much of Kent had been occupied by a people, for whom we have no name, who used the hill-letter m but that at dates prior to 130BC northwest Kent was settled by the Canti and northeast Kent by a people who used the hill-letter r. It has been shown that the Canti expanded north and west to take over London, Wandlebury hillfort (southeast of Cambridge) and Bicester (Oxfordshire), this movement taking place earlier than 130 BC. At some point the Canti living north of the Thames were given a new name, the Catuvellauni. The expansion of the Catuvellauni displaced some Trinovantes northwards into Lincolnshire. At a later date, after 120BC, the Catuvellauni moved east into Essex, displacing some Trinovantes into north and east Kent. This movement of the Trinovantes caused some m-people to move west to the Hastings area and the upper reaches of the East Rother river. It also caused some r-people to move from northeast Kent to the area around Mount Caburn in East Sussex. These points are discussed in detail in Chapter 23, 2.6 and 2.6.1.

9             By studying the movement of tribes within southeast England in association with information as to the find-spots of coins it has been possible to assign the Kentish Primary Series potin coins to the m-people living in Kent, the Flat Linear I potin coins to the r-people of northeast Kent and the Flat Linear II potin coins to the Catuvellauni. In addition, if the Trinovantes were using gold coins prior to the displacement of some of them to Lincolnshire (prior to 130BC) then one should expect to find some Trinovantian gold coins in Lincolnshire. But the Trinovantes who moved to Lincolnshire can reasonably be expected to have maintained contact with the Trinovantes still living in east Essex, so it is reasonable to expect to find in Lincolnshire Trinovantian coins minted after 130BC at the Trinovantian capital, Camulo at Colchester.  Moreover, since Trinovantes took control of southeast Kent around 100BC one should expect to find in Kent Trinovantian coins minted after that date. But since it is reasonable to expect that when the Catuvellauni  took control of the Trinovantian state by capturing Camulo at Colchester they also acquired control over the Trinovantes living in southeast Kent, then one should expect to see Catuvellaunian coins superseding Trinovantian issues in southeast Kent, presumably at some time in the second half of the first century BC.

10           It has also been possible to identify the Celtic names of some fifty hill-forts, some with certainty, others with a little less certainty. Amongst those where we can be certain are Averno (part of the name Duroaverno) at Bigbury Camp, just west of Canterbury, Anicetis  just north of Ansty in Wiltshire, Nedionemedon (Medionemeton in Ravenna) just northeast of Bar Hill in East Dunbartonshire, Magnis at Credenhill Park Wood, just northeast of Kenchester in Herefordshire, Bograndium  at Braco in Perth and Kinross, Magno (part of the name Noviomagno in Ravenna) at Maiden Castle in Dorset and Vresmedenaci (Bresnetenaci in Ravenna) at Portfield Camp, southeast of Whalley in Lancashire. Full details are given in “The Celtic names of some British hillforts”, accessible from the main menu above. But, as noted, in some cases we cannot be quite certain, as, for example, in the case of Cermium, apparently the hillfort at Castle Law, Forgandenny in Perth and Kinross. In this example, as in some others, there are two hillforts near the location of a place named in Ravenna, but since the name could apply to either hillfort we could only be certain as to which should be identified as the Celtic predecessor of the Roman fort if the archaeologists could tell us that one of the hillforts was certainly occupied when the Romans reached the area concerned, whereas the other certainly was not.  

11           A number of other points have arisen in the course of this study, some of them worth mentioning in this summary.

11.1       There are ten names in Ravenna which have an essa-type ending. In all of these names the first part of the name includes either the hill-letter s, as in Camulosessa, Certisnassa and Racstomessa (Raxtomessa) or the corresponding river-letter b, as in Abisson, Devionisso (bv) and Duabsissis (the first s in this name is the river-letter s, not the hill-letter s). It is thus quite clear that the essa-ending was used by the people who employed the hill-letter s. The writer is not aware of any evidence that any other settlers used this ending. The essa-ending is discussed in Chapter 4.         

11.2       The Ravenna names from Serduno to Maia Fanococidi have long been held to be the names of forts on Hadrian’s Wall (though some scholars have regarded Fanococidi as being the fort at Bewcastle), but in fact they predate Hadrian’s Wall.  They appear to refer to forts forming part of the Trajanic frontier. The frontier ran west from Serduno at Wallsend to Condecor at Benwell, with no fort at Newcastle. It then crossed the country a little to the south of the line later taken by the Wall, all the way to Fanococidi at Kirkbride.  After Benwell the frontier ran west to Vindovala, presumably just south of Rudchester, and then on to Onno, most probably in the vicinity of the village of Halton. From there the frontier ran west to Celunno (a development of the Flavian name Celovion), most probably close to the point where the Stanegate crossed the North Tyne. From there the frontier ran west to Brocoliti at Newbrough, Velurcion at Grindon Hill and Esica at Haltwhistle Burn. Magnis at Carvoran would be the next fort on the frontier, though it is not mentioned at this point in Ravenna since it was listed earlier in a group of Flavian forts. The Trajanic frontier then ran on to Banna at Throp and over to Cambaglanda at Birdoswald, though this fort is not mentioned at this point in Ravenna since it, too, was mentioned earlier in the group of Flavian forts. The frontier then ran west to Uxelludamo, apparently initially at Old Church Brampton and later, with the name changed slightly to Uxelodunum, at Castlesteads, and from there on to Lagubalium at Carlisle. This fort, too, was listed earlier in that group of Flavian forts and so is not mentioned at this point in Ravenna. West of Carlisle the Trajanic frontier ran to Avalava at Burgh-by-Sands III and then on to Fanococidi at Kirkbride via Maia up on Fingland Ridge. There was of course no frontier wall at that time, just a string of forts stretching from one side of the country to the other. These points are discussed in more detail in Chapter 20, 3.

11.3       The Ilam pan (also known as the Staffordshire Moorlands pan) has in the past been held to be a souvenir of Hadrian’s Wall. However, whilst the vessel itself may well be old the inscription on it is modern. The odd text RIGOREVALIAELIDRACONIS is in fact an anagram, the original text apparently being LEONARD RIVER ISO GALICIA. There is indeed a river Iso in Galicia, but the word RIVER can also be taken together with LEONARD to form a personal name, LEONARD RIVER. This is a thinly veiled reference to A.L.F. Rivet, apparently known to his friends as Leo Rivet. ‘Leo’ is derived from Leonard and the modern English word ‘rivet’ is apparently derived from Old French ‘river’. The pan with its inscription is perhaps most likely to be a university prank, the findspot of the pan being only some 16 miles from the university where Rivet worked for many years. This does not of course mean that Rivet himself had anything to do with the prank. The inscription may have been carried out by admirers after Rivet’s death. The pan is discussed in detail in Chapter 9.

11.4       The Rudge cup and Amiens patera have long been considered to be souvenirs of Hadrian’s Wall, but they appear in fact to date from an earlier period. They appear to commemorate the building of new or recommissioned forts to form part of the Trajanic frontier from the Tyne to the Solway. The lists of forts on the pan and patera start in the west. They do not include Fanococidi at Kirkbride because the fort there was already operational and was merely incorporated in the frontier as its westernmost fort. Both pan and patera start their list with Mais up on Fingland Ridge and continue with Aballava at Burgh-by-Sands III. They do not list Carlisle because the fort there was fully operational and was merely incorporated in the new frontier system. Both then include Uxelodunum (given as Uxelodum on the cup) at Castlesteads and then Cambaglanda at Birdoswald (rendered Camboglans on the cup and Cambog…s on the patera). Both then include the Trajanic fortlet Banna at Throp and the patera then lists Esica at Haltwhistle Burn. The above points are discussed in detail in Chapters 8 and 20, 3.

11.5       The Ravenna names from Velunia to Credigone do not actually represent forts on the Antonine Wall but fortlets built as part of the preparations for the construction of that Wall. The fortlets are evenly spaced 5.6 kilometres apart (correct to the first decimal place) from Velunia at Carriden to Credigone at Duntocher, except for the stretch from Mumrills to Camelon (which was included in the fortlet system) which was longer, at 5.9 kilometres. The fortlets appear to have been located on the line the Wall was originally intended to follow, though changes were made to the line between Mumrills and Seabegs and also to the west of Cadder. It seems most likely that the fortlets were occupied by teams of surveyors charged with the task of finding the best line for the Wall, each team surveying the area between its own fortlet and the next fortlet to the west. The team in Duntocher therefore had to find the best line for the Wall between Duntocher and the river Clyde. These points are discussed in detail in Chapter 22.

12        The present writer is fully aware that older scholars in the field of Romano-British place-name studies, especially perhaps those who have taught the subject in universities for many years, will find it difficult to accept the ideas put forward on this website, if only because they are so radically different from conventional wisdom on this subject. But the attentive reader will have noted that the information in Chapters 1 and 2 as to the building-blocks of Romano-British place-names, and as to how those building blocks were combined to create compound place-names, has been applied in a consistent and logical manner to place-names throughout Great Britain. Many names which have been explained in the past by scholars have been explained in a new and more consistent manner, and many names which scholars have dismissed as being hopelessly corrupt, names such as Omiretedertis, Subdobiadon, Ugrulentum and  Ugueste, have also been explained. These names are not hopelessly corrupt – it has just not been possible to explain them using the conventional methodology. But if one adopts the principles set out in Chapters 1 and 2 of this website then the above names take their place alongside all the other topographical names in Ravenna and Ptolemy and their explanation is in fact relatively straightforward.

13        The river-names in Ravenna and Ptolemy are discussed in detail in Chapter 19, where there are links to notes relating to selected one of the names. The correspondence of land-names (place-names comprising one or more hill-letters) and river-names has been explained in the case of the place-name Bresnetenaci/Bremetonnaci and the river-name Belisama, but there are many other examples, such as:

a)         Camulodono at Skipton in Yorkshire stood on the Eller Beck close to its confluence with the river Aire. The river-letter r in Eller and Aire corresponds to the hill-letter m in Camulodono. The river-letter corresponding to the hill-letter n of the place-name is l for a minor river, this l being present in Eller, and m for a main river. On the face of it, then, the Celtic name of the Aire probably included an element such as Mor. Now, the ND refers to a fort called Morbio to the north of Doncaster and so it seems likely that Morbio was in fact the then name of the river Aire, the river-name being transferred to a late-Roman fort at Castleford;

b)        Elconionemedo at or near Launceston stood on or near a river called the Tamaris (now the Tamar). The river-name includes the river-letters t, m and r corresponding to the hill-letters l, n and m of the place-name;

c)        Lectoseto (later Lectoceto) at Towcester stood on or near a river now called the Tove. The river-name will earlier have been Tobe, having the river-letters t and b corresponding to the hill-letters l and s of the place-name. The river-letter b has simply changed to v in the modern river-name;

d)        Londinion (later Londinium) stood on a river called the Tamesis (now the Thames). The river-name Tamesis includes the river-letters t and m corresponding to the hill-letters l and n in the place-name;

e)       Lucamosessa (later Camulosessa) at Malton in Yorkshire stood on a river once called the Terbentio (later Derventio, now the Derwent). The river-letters t, r and b of the Terb part of the river-name correspond to the hill-letters l1, m and s of the place-name. The en of the river-name was just a name-ending before the final t was added by people who used the hill-letter l2; and

f)      Rugulentum (later Ugrulentum), at Cardean in Angus, stood at the point where two headwaters now called the Isla and Dean came together to form the river earlier called the Matovion (today it also has the name Isla). The river-name Matovion includes the river-letters m and t corresponding to the hill-letters n and l of the place-name, the t of Matovion coming after the m because the hill-letter l in Rugulentum is l2 and the l2-people placed their river-letter t at the end of the existing river-name. In addition the s and l of the river-name Isla correspond to the hill-letters r and n of Rugulentum. The d in the river-name Dean Water will just be a modified river-letter t corresponding to the hill-letter l of Rugulentum.

13.1     Then again, by working backwards from the known river-name Vedra (the Wear) it has been possible to show that the Celtic name of the fort at Chester-le-Street was most probably Mucoganges or Lucoganges, so that when explaining the derivation of the name Coganges/Concangios there was no sense whatsoever in looking for any modern Celtic words resembling Coganges or Concangios. Likewise, by working backwards from the known river-name Bdora/Boderia (the Forth) it has been possible to show that the pre-Roman name of Stirling was most probably Mugulesde (Ugueste in Ravenna).  In addition, the river-name Dart (in Devon) indicates that the Celtic predecessor of Ravenna’s Melamoni was called Lelamon.

14        To press home the point that the conventional approach to explaining Romano-British place-names is simply misguided, attention is drawn to four groups of Romano-British place-names.

14.1     The first group comprises compound names including the old-style element nd meaning ‘hill summit’ and including the hill-letter n. These names are:










































uesedo (AI)







(No justification is given here for the restored forms of the names shown in italics. These names are all explained on other pages of this website. Any reader wishing to check up on any particular name is invited to use the Search function available at the top right-hand corner of each page of this website). Leaving aside place-names in which the nd element, including the hill-letter n, is not qualified, as in Anderelion, or is qualified solely by a b meaning ‘high’ and/or a c meaning ‘steep’, so far as the present writer is aware there are no names in Ravenna and Ptolemy including the element nd other than those listed above. It is quite clear that we are dealing here with a family of topographical compounds in which the element nd, including the hill-letter n and meaning ‘hill summit‘, is compounded with one of the other hill-letters. In some names the other hill-letter is itself unqualified, but in Clindum the other hill-letter l is qualified by c meaning 'steep', in Gambaglanda the other hill-letter l is qualified by bag meaning ‘high, steep’ and in Bograndium the other hill-letter r is qualified by bog, also meaning ‘high, steep’. There is therefore no sense in looking around for modern Celtic words resembling any of those compounds and then inventing hypothetical roots from which the place-names concerned might be assumed to be derived. Thus, if the invention of the hypothetical roots lindo, thought to mean ‘pool, lake’, and mandu, thought to mean ‘small horse, pony’, was based solely on the place-name evidence then the time has surely now come to discard those hypothetical roots.

14.2     The second group of names comprises those names including a transitional element. These names are:







































io (AI)







So far as can be seen there are no other names of this form in Ravenna and Ptolemy, that is to say a name having one consonant preceded by a b and followed by a c (or g), or preceded by a c (or g) and followed by a t. And all of the names in the above list include one of the hill-letters in the middle column, i.e. they all include an element called a transitional element on this website. One can see above that there are examples of transitional elements in all hill-letters except m. This does not mean that there never was a transitional element in the hill-letter m – it merely means that we don’t have an example of such an element in Ravenna and Ptolemy. But again we are looking at a family of topographical elements, using the different hill-letters, so there is no sense in looking for modern Celtic words resembling any of the above transitional elements and then inventing hypothetical roots to explain the development of the Romano-British names. But in the past the hypothetical root qert, taken to mean ‘to turn or twist’ or ‘winding one’, was invented to explain the cert of Certisnassa, the hypothetical root bolgo, taken to mean ‘bag, bulge’, to explain the bulg of Bladobulgio (Blatobulgio) and the hypothetical root brocco, taken to mean ‘badger’ or ‘pointed’, to explain the Broc of Brocara (Brocavo in the AI) and Brocoliti. Thus, if the invention of qert, bolgo and brocco was based solely on place-name evidence then it is now time to dispense with those hypothetical roots.

14.3     The third group of names comprises those names including an inversion-type element meaning, in modern English, ‘a steep, high hill’ or ‘a high, steep hill’. These names are:




























So far as can be seen there are no other names in Ravenna and Ptolemy which include a consonant followed by a c (or g) and a t.  And every name in the above list includes in the middle column an inversion-type element using one of the hill-letters. We have examples of this element in all hill-letters except s. It would appear, however, that Racatonium was earlier Celtic Rascatonion, the asc element surviving in the river-name North Esk. The scat element is an inversion-type element of the kind discussed here. In addition, another example is almost certainly to be seen in the stuc element of Ptolemy's river-name Stuccia, probably originally Stuccla, which is a land-name transferred to a river. Two other names should apparently be included in the list, namely the Viguto of Duroviguto and Itucodon. It is suggested elsewhere on this website that Viguto was originally Virguto, though the hill-letter need not have been an r, and that Itucodon was originally Litucodon, though the initial letter need not have been an l. But again it is clear that we are dealing with a family of topographical elements using the different hill-letters, so there is no need to look around for modern Celtic words resembling any of those elements and then to invent hypothetical roots to explain the derivation of the Romano-British names.

14.4     Finally we may consider the banva- and venta-type names. These are shown below. The list includes banva-type names which use the letter c meaning ‘steep’ rather than the letter b meaning ‘high’.


                            Names of banva-type                                                     Names of venta-type















































































We have no example of a banva-type name using the hill-letter s and no example of a venta-type name using the hill-letter m or s. Again, this does not mean that such elements never existed – it merely means that we do not now have examples of them in Ravenna and Ptolemy. But yet again it is clear that we are dealing with a family of topographical elements using the different hill-letters, so again there is no need to invent hypothetical roots to explain the Romano-British place-names. It follows that if the invention of the hypothetical roots banno (or banna), taken to mean ‘peak, horn’ (and thought in the past to explain names such as Banna) and venta, taken to mean ‘field’ or ‘market’ (and thought in the past to explain the name Venta) was based solely on the place-name evidence then we should now discard those hypothetical roots.

15     It should thus now be clear, even to the most dyed-in-the-wool Celtic academic, that the current technique for explaining Romano-British place-names, which has held sway for very many years and was discussed briefly in the Introduction to this study, is simply misguided. The Celts did not coin place-names by using words with such unlikely meanings as ‘wagon-ford, chariot-ford’, ‘horse-people’, ‘apple orchard’, ‘goat path’, ‘place on sword river’, ‘small horse or pony’, ‘bear place’ or ‘cloud, mass of clouds’ (all examples quoted in Rivet and Smith 1979). Instead they coined simple topographical names by using one hill-letter with a name-ending and where the hill-letter may or may not be qualified by a letter b or t meaning ‘high’ and/or a letter c or g meaning ‘steep’, or where the hill-letter is itself used to qualify a v meaning ‘on the slope’, or ‘on the side of’, or a d meaning ‘summit’. Or they coined compound topographical place-names comprising two or more hill-letters, where each successive wave of settlers added its own element, comprising its own hill-letter, to an already existing name. And in the case of river-names the Celts did not coin names using words with such fancy meanings as ‘winding one’, ‘the washer’, ‘vigorous stream’, ‘babbling river’ or ‘old river’ (again all examples quoted in Rivet and Smith 1979). Instead they formed simple river names by using a river-letter accompanied by a name-ending, or compound river-names where each successive wave of settlers added its own river-letter to an already existing river-name. 




[This page was last modified on 28 March 2022]