Celtic place- and river-names and Iron Age coins
1.1 The present chapter is primarily concerned with the potin coins used in southeast England in the 2nd and 1st centuries BC, but reference will be made here and there to other Iron Age coins.
1.2 The term ‘Celtic’ is used broadly below to refer to all of the tribes or other groups who lived in Britain prior to the Roman invasion in AD43. The term ‘Iron Age’ is applied to the coinage used by those Celtic tribes since the use of that term is now very well established in the literature in this field of study and there would appear to be no advantage in replacing ‘Iron Age’ by ‘Celtic’ or any other term.
1.3 Section 2 below discusses the different Celtic tribes who inhabited southeast and eastern England in the 2nd and 1st centuries BC and how some of those tribes migrated from one area to another. To understand Section 2 it is necessary to be familiar with the structure of Celtic place-names and river-names as explained on other pages of this website. The structure of the place-names is explained in chapters 1 and 2, that of the river-names in chapter 19. A brief summary is included below for the benefit of those readers who have not read those earlier chapters.
1.3.1 At different times prior to the Roman invasion of Britain in AD43 different Celtic tribes migrated from the Continent to Britain. When coining topographical place-names those different tribes used different consonants of the alphabet to refer to ‘a hill’. They also used different consonants to refer to ‘a river’. Five different groups can be identified. One group used the hill-letter s and the river-letter b. A second group used the hill-letter m and the river-letter r. A third group used the hill-letter r and the river-letter s. A fourth group used the hill-letter l and the river-letter t. The fifth group used the hill-letter n and the river-letters m and l. The fifth group appears to have applied the river-letter m to major rivers and the river-letter l to minor rivers, where a minor river may be a small river which flows directly into the sea or it may be a tributary or headwater of another river. The picture is somewhat muddied by the fact that each of the hill-letters n and l appears to have been used by two different groups who arrived in Britain at different times. The normal chronological order of the hill-letters for Britain as a whole appears to be n1, s, m, r, l1, n2, l2, though there is some regional variation, as in Lincolnshire, where n2 is earlier than l1, and central Scotland, where r is earlier than m. In addition, the hill-letters in a place-name may be qualified by a b or t meaning ‘high’ or a c or g meaning ‘steep’. Some place-names are termed ‘old-style names’ on this website, which means that the qualifier comes before the hill-letter in the name. Other place-names are what is here called ‘inversion-type’. In these names the qualifier comes after the hill-letter in the name. A few place-names are transitional. In these names a hill-letter has two qualifiers, one coming before the hill-letter, the other after. In old-style place-names, in which the qualifier comes before the hill-letter, it is the letter b which is used to signify ‘high’. In inversion-type names, in which the qualifier comes after the hill-letter, it is the letter t which represents ‘high’. This is also true for transitional names if the qualifier ‘high’ comes after the hill-letter. Lastly, the letters d and v are respectively used to represent the summit of a hill and a slope, a hillside. The letters d and v are used as nouns within place-names, being qualified by one of the hill-letters.
The above is a little indigestible, but a few examples should illustrate the points made. As an example of an old-style place-name in the hill-letter r we can take the name Bereda. In this name Ber means ‘high hill’ and Bered means ‘high hill summit’. As an example of a compound old-style name we can look at Clindum. Cl means ‘steep hill’ and the element Cl qualifies the later element ind which means ‘hill summit’. The name as a whole thus refers to a place on the summit of a steep hill and it tells us that the place concerned was originally occupied by people who used the hill-letter l1 but was taken over by people who used the hill-letter n2. A similar example is Bograndium. The Bogr element means ‘high steep hill’ and qualifies the later element and which means ‘hill summit’. The name refers to a place on the summit of a high, steep hill and indicates that the place concerned was first occupied by people who used the hill-letter r but was taken over by people who used the hill-letter n2. As an example of the use of the letter v meaning ‘slope’ we can take the name Bamvocalia (the Celtic form of Ravenna’s Pampocalia). Bam means ‘high hill’ and Bamv means ‘high hill slope’. The element Bamv qualifies the later element cal which means ‘steep hill’. The name as a whole thus refers to a place on the slope, the side, of a high, steep hill and indicates that the place concerned was once occupied by people who used the hill-letter m but was later taken over by people who used the hill-letter l1. The b/v/p interchange is fairly common in Celtic/Romano-British place-names, so it need cause no surprise that Bamv could change to Pamp. As an example of a simple transitional name we can take Cunetio, in which the hill-letter used is n2 and is qualified by c meaning ‘steep’ and t meaning ‘high’. As an example of a simple inversion-type name we can consider Macatonion, in which the hill-letter is m and is qualified by c meaning ‘steep’ and t meaning ‘high’. Finally, the somewhat more complicated name Marcotacson (the Celtic form of Ravenna’s Marcotaxon). The earliest element is the old-style cs which means ‘steep hill’. This qualifies the inversion-type element arcot, which means ‘hill steep high’. And arcotacs qualifies the hill-letter m used in an inversion-type manner. This name tells us that the place concerned was once occupied by people who used the hill-letter s but was taken over by people who used the hill-letter r. Later still it was taken over by people who used the hill-letter m. These few examples are merely intended to illustrate the points raised in the preceding paragraph in order to help the reader understand more easily the explanation presented below in section 2 relating to tribal migration. For a detailed discussion of the names given above and their identification the reader should consult the Alphabetical List.
1.3.2 It is also possible to identify which tribes in east and southeast England used which hill-letters, as explained on the page “Ptolemy’s Celtic tribes in Britain” of this website. Briefly, and confining our attention to east and southeast England,
the Iceni used the hill-letter r
the Trinovantes and Coritani used the hill-letter l (actually l1)
the Canti and the Catuvellauni used the hill-letter n (actually n2).
Note that the tribal names are those given by Ptolemy, though it is not altogether clear that those names were actually in use in the 2nd and 1st centuries BC, at least in those forms. It has become fashionable in recent years, for example, to prefer the form Corieltauvi to Ptolemy’s Coritani.
1.3.3 By studying the place-names and the chronological order of the hill-letters in them it is possible to identify the different areas which were settled by the different tribes when they arrived from the Continent, and possible also to see how subsequent expansion of one tribe caused other tribes to move from one area to another in a kind of domino effect. This is important for identifying who manufactured the different series of potin coins, when production started and when it stopped. These points are discussed below.
2 Tribal migration
2.1 For the purposes of the present chapter it is sufficient to start with the people who used the hill-letter m, though there were of course earlier Celtic settlers, namely those who used the hill-letters n1 and s. We have no name for those m-people, but at one time they occupied southeast Kent. The only place-name we have is Lemanis, which was the name of the Roman fort at Lympne, though the name will have been transferred to the fort from a Celtic settlement. That settlement may have been on the same site as the fort, but it may equally well have been somewhere else in the vicinity. But it will have been close to the river East Rother, the final r of this river-name being the river-letter corresponding to the hill-letter m. We also see the river-letter r in the river-name Durbis, now the Dour, the river at Dover, and again in the river-name Turupis, now the Stour. But we also see evidence of the m-people in north Kent, for the river-letter r is present in the river-name Darent. In addition, the place-name Noviomago of Iter II of the Antonine Itinerary is generally held to refer to the Romano-British settlement at Crayford, but the element mago is most likely to have been Celtic magno as in the case of Ravenna’s Navimag(n)o, where the earlier Celtic name Magno was that of the hillfort called The Trundle, a little to the north of Chichester. One sees the Magno element again in Ravenna’s Noviomagno, where Celtic Magno was the Maiden Castle hillfort in Dorset. And one sees the element as Magnis in Ravenna and in Iter XII of the Antonine Itinerary. In Ravenna and the AI the name refers to the Romano-British town at Kenchester in Herefordshire, but the earlier Celtic Magnis will have been the hillfort a little to the northeast, in Credenhill Park Wood. To return to the Noviomago of Iter II, the earlier Celtic Magno will have been a hillfort adjacent a steep hill, most probably at the top of it as in the other examples just mentioned, somewhere not too far from Crayford, and that hillfort may have been east or west of the Darent. And of course the initial M of Magno is the hill-letter m. We can thus safely conclude that m-people lived in north Kent in the region around the river Darent.
2.1.1 But we see evidence of m-people also outside Kent. Magno at Maiden Castle in Dorset and Mag(n)o at the Trundle in West Sussex were both mentioned above in paragraph 2.1. But we see the hill-letter m also in Verlamion at Prae Wood, just west of St. Albans, in Camuloduno at Colchester, in Combredovio (the Celtic form of Conbretovio in Iter IX of the AI) in the Coddenham/Baylham House area, in Camborico (in Iter V of the Antonine Itinerary) apparently at Thetford and in Gambrand, apparently the Celtic name of South Ferriby in north Lincolnshire (or of a place up on the high ground immediately southeast of South Ferriby).
2.2 The first migrants to arrive from the Continent after the m-people were the people who used the hill-letter r. Some of those people occupied northeast Kent, east of the river Medway (Brabis, the Celtic predecessor of Durobrabis at Rochester, is the only old-style place-name in the hill-letter r to survive in that area), others settled in Norfolk, and at some point r-people took over South Ferriby (Gambrand) in North Lincolnshire and founded Ancaster (Croco(n)calana) in south Lincolnshire. There were also r-people in East Sussex, though the only place-name we have from that area is Racstomessa (Ravenna’s Raxtomessa) at Mount Caburn, but since the Rac element is inversion-type the name is later in date than the r-elements in Gambrand and Croco(n)calana. We can’t therefore be sure whether the r-people in East Sussex arrived directly from the Continent or were migrants from some other area of Britain, such as northeast Kent (but see paragraph 2.6.1 below).
2.3 The next migrants to consider are those who used the hill-letter l1. Those people settled in the area north of the Thames estuary, where we see the old-style names Londinion (London), Camulo (Colchester) and probably Libonde (Wandlebury, southeast of Cambridge). We perhaps see the corresponding river-letter t used in the old-style manner in the river-name Blyth (t → th). Those l1-people north of the Thames estuary were known, or came to be known, as the Trinovantes.
2.4 The next migrants were those who used the hill-letter n2. A first group, the Atrebates (this name is used broadly here to include the Regni), landed on the coast of south Hampshire and West Sussex. They gained control of Magno at The Trundle, of Claducendum (the Clausentum of Iter VII of the Antonine Itinerary) at Old Winchester Hill and of Magno at Maiden Castle. A second group, the Canti (early Catuvellauni), took control of northwest Kent, west of the river Medway. We see the river-letter m, corresponding to the hill-letter n, in the Med element of the river-name Medway. We also have the name Vagniacis of the Romano-British religious centre at Springhead, this name having been transferred from a nearby Celtic settlement, the first element of the Celtic name most probably having been the old-style element Bagn meaning ‘high steep hill’. And we have the old-style name Magno, discussed above in paragraph 2.1, of a hillfort somewhere in the vicinity of Crayford.
2.4.1 There appears to have been a quite separate landing of n2 -people in Lincolnshire, where they took over Ancaster (Croco(n)calana) and South Ferriby (Gambrand) and founded Caistor (Banvobalum, the Celtic form of the Bannovalum of Ravenna). They also appear to have founded Lincoln (Lindum, the initial L being inversion-type).
2.5 Finally we have to deal with the l2-people, who appear to have settled in the area between the rivers Medway and Darent. The d in Medway is most probably the river-letter t, changed to d, corresponding to the hill-letter l2. The t in Darent is also the river-letter corresponding to the hill-letter l2. There is also a river Eden, which is a tributary of the Medway. The d of this name is most probably also the river-letter t, changed to d, corresponding to the hill-letter l2.
2.6 Paragraphs 2.1 to 2.5 identify the different tribes who migrated to southeast England from the Continent, but now we have to consider how some of those tribes migrated within that region of Britain. The first move was made by the Canti, who expanded north of the Thames, taking London from the Trinovantes (they added the nd element so that the name became Londinion). Note that Ptolemy does in fact assign Londinium to the Canti, but for convenience hereinafter the term Canti will be applied only to those Canti who remained in northwest Kent. Those who lived north of the Thames will be called the Catuvellauni, the name by which they later came to be known. The Catuvellauni moved on north along the valley of the river Lee, the L of which is the river-letter (for minor rivers and tributaries) corresponding to the hill-letter n. They appear to have moved further north and to have seized control of Wandlebury hillfort, this now becoming Libonde. The important point to note is that this northward movement of the Catuvellauni displaced some Trinovantes, who migrated north to Lincolnshire and seized Ancaster (Croco(n)calana), Caistor (Banvobalum) and Marton (Segeloci). The Catuvellauni also expanded westwards from the London-Wandlebury axis since we see Brinavis at Bicester in Oxfordshire. At a somewhat later date the Catuvellauni moved east of the river Lee and displaced more Trinovantes, these then migrating to Kent. By this time the Trinovantes were coining inversion-type names, which is why we see the Celtic place-name Lemanis and the river-name Rother (the L is used in an inversion-type manner in the place-name and the corresponding river-letter t, changed to th, comes before the final r of the river-name). We also see the river-letter t coming before the r in the river-names Stour, Durbis (the Dour at Dover, the t being changed to d) and Darent (again the t being changed to d). It is thus clear that the Trinovantes moved from Essex to Kent at a time after they had changed over to using inversion-type names. The only two names which might appear inconsistent with this are Regulbio (at Reculver) and Levo (the Celtic predecessor of Durolevo in the Sittingbourne area), but in Regulbio the l may well have been applied at the time of that migration, this l then qualifying the later inversion-type element Reg, which means ‘hill steep’. The name Levo is not clear, but if the L is the hill-letter l, then this name, too, might well have been created at the time of the migration from Essex into Kent. Note that this migration of Trinovantes into Kent displaced some of the people who used the hill-letter m. We thus see Ravenna’s Mutuantonis, where Mut is an inversion-type element meaning ‘hill high’, apparently at Hastings. In addition the initial R of Rother was clearly applied by m-people who had switched to using inversion-type names. Apparently m-people had moved further upstream from Celtic Lemanis.
2.6.1 Note that when some of the Trinovantes crossed over to east Kent it is probable that they displaced some of the r-people living in northeast Kent, these r-people, by now using inversion-type names, then moving to East Sussex to become the r-people living around Racstomessa at Mount Caburn (discussed above in paragraph 2.2).
2.7 Paragraph 2.6 sets out the main population movement within southeast and eastern England in the 2nd and 1st centuries BC, but there are two minor movements to take into account before we find the population distribution which appears to have existed just before Caesar’s arrival in 55BC. These minor movements concern Canterbury and St Albans.
2.7.1 As to Canterbury, the river-letters t and r in the river-name Stour indicate that the area around the river was at one time occupied by people who used the hill-letter m but was at some point taken over by people who used the hill-letter l1, i.e. by the Trinovantes. It is possible that the settlement on St. Thomas’s Hill in the northern outskirts of modern Canterbury (Clark and Lane 2015, 1) had a name including the hill-letter m, but the name is unlikely also to have included the hill-letter l for reasons set out in paragraph 3.4 below. However, the Trinovantes must have settled somewhere along the course of the river Stour to account for the river-letter t in the name Stour. It is probable that Bigbury Camp went by the same name as the settlement on St. Thomas’s Hill, this implying that the people lived and worked in the settlement on St. Thomas’s Hill and only used Bigbury Camp as a place of refuge in times of danger. It is believed that the settlement on St. Thomas’s Hill was abandoned around 100BC (Clark and Lane 2015, 6), so this is presumably also true of Bigbury Camp. We learn from Commentarii de bello gallico V,9 that Caesar marched about 12 miles from the coast and reached a river where the Britons were assembled with their cavalry and chariots. Apparently the British forces were pushed back by the Roman cavalry and sought refuge in a place well protected by natural features and by man-made defences. It was remiss of Caesar not to give us the name of the place, but the description and the location appear entirely appropriate for Bigbury Camp. It would appear, then, that Bigbury Camp had been abandoned by its earlier occupants sometime around 100BC but was later re-occupied, this time by the r-people, who gave it the name Averno, where the ver element means ‘slope/side of a hill’. That Bigbury Camp had been abandoned is supported by the place-name Averno itself, since normally when newcomers took over a settlement they simply added their own place-name element to the already existing place-name. Thus when the Catuvellauni took over Colchester they simply added their element duno to the Trinovantian name Camulo. But in the case of Bigbury Camp the newcomers coined a wholly new name, Averno, this implying that there was no settlement at Bigbury Camp when the r-people moved in.
2.7.2 In the case of St Albans it would appear that north of the Thames the Catuvellauni exerted pressure westwards as well as eastwards, since about the same time that the eastward expansion of the Catuvellauni displaced some Trinovantes towards Kent, other Trinovantes seized Prae Wood from the m-people, the l in Verlamion being used in an inversion-type manner. But the Trinovantes appear not to have held Prae Wood for very long, since the hillfort was taken over by r-people, presumably from the area around Dunstable, where we see Cobrivis, the Celtic predecessor of Roman Durocobrivis, which was located at Dunstable itself. But the control of the r-people over Prae Wood was probably short-lived since when the Catuvellauni established their oppidum at Wheathampstead it is unlikely that they would tolerate the presence of a potentially hostile hillfort only a few kilometres away at Prae Wood. It would thus appear that at some date before Caesar’s arrival in 55BC the Catuvellauni moved the inhabitants of Prae Wood into a new settlement on the lower slopes of the hill, close to the river Ver, this new settlement being given the name of the old, i.e. Verlamion.
2.8 The above paragraphs set out what appears to have been the distribution of the tribes in southeast England on the eve of Caesar’s invasion in 55BC. At the end of hostilities in 54BC Caesar demanded hostages and payment of an annual tribute from Cassivellaunus and ordered him to desist from any action against the Trinovantes and their leader, Mandubracius. There were thus two tribes north of the Thames, the Trinovantes in eastern Essex and the Catuvellauni further west. South of the Thames there will have been Canti in the area west of the Darent, l2-people between the Darent and the Medway, r-people east of the Medway, including Rochester, Reculver and Bigbury Camp, and Trinovantes in southeast Kent. The rulers of those four areas in Kent are presumably the four kings in Kent mentioned by Caesar, though the ruler of the area west of the Darent was presumably a sub-king under the overlordship of Cassivellaunus, king of the Catuvellauni, and the ruler of southeast Kent was presumably a sub-king under the overlordship of the leader of the Trinovantes.
2.9 There were of course political changes within southeast England after Caesar returned to the Continent.
2.9.1 It would appear that the Catuvellaunian elite abandoned their oppidum at Wheathampstead and chose the new settlement down below Prae Wood (see paragraph 2.7.2) as their new base. The various Iron Age ditches in the vicinity of St. Albans presumably date back to this period. It was this Catuvellaunian Verlamion which was later romanised to become Romano-British Verulamium.
2.9.2 As regards Canterbury, whilst there was apparently no settlement on the low ground by the river at the time of Caesar’s invasions of 55 and 54BC, excavation has shown that there most certainly was a settlement on the low ground before the Roman invasion of 43AD. It is believed that Bigbury Camp was finally abandoned after Caesar’s invasion. Moreover the Romano-British name Duroaverno indicates that there had been a Celtic settlement called Averno before the Roman invasion of AD43. It is thus reasonably clear that the new settlement founded down on the low ground by the river Stour was also called Averno. But it will not have been founded by the r-people themselves since they are hardly likely to have applied to a settlement on low-lying, flat ground a name which actually refers to the slope, the side of a hill. It follows that someone else moved the inhabitants of Bigbury Camp down to that new settlement on the low ground by the river Stour and they gave that new settlement the name, Averno, of Bigbury Camp. Bearing in mind that the balance of power had swung in favour of the Catuvellauni - it was after all Cassivellaunus of the Catuvellauni who led the British forces against Caesar - it seems most likely that it was Cassivellaunus who moved the inhabitants of Bigbury Camp down the hill to the new settlement. From that point on the people of the Canterbury area, and presumably all of the r-people in northeast Kent, will have been ruled by the Catuvellauni.
2.9.3 But arguably the biggest change in southeast England was that at some point, most probably after the death of Caesar, the Catuvellauni did indeed gain control over the Trinovantes north of the Thames, so Trinovantian Camulo at Colchester became Catuvellaunian Camuloduno. Whether gaining control over the Trinovantes north of the Thames also automatically and simultaneously gave the Catuvellaunian king control over the Trinovantes in southeastern Kent is not clear, but it probably did, if he had not already absorbed them into his kingdom.
2.10 A point not discussed above is the date of arrival in Britain of the various tribes or groups who used the hill-letters r, l1, n2 and l2. It would appear that at least those who used the hill-letters r, l1 and n2 were all in Britain by 150 BC at the latest. The reason for this is that the above hill-letters were used by tribes in widely different parts of Britain (as explained on the page “Ptolemy’s Celtic tribes in Britain” of this website), but it was only those in the east, southeast, south and southwest of England who adopted coinage. For example, the hill-letter n2 was used not only by the Catuvellauni and Atrebates but also by the Brigantes, as shown by the place-names Cambrolanda (Ravenna’s Cambroianna) and Carbandium (Ravenna’s Carbantium), and also by groups who settled still further north, as shown by the names Binovia (the Vinovia of Ravenna and the AI), Bindolande (Ravenna’s Vindolande), Magnis and Cambaglanda (the Gabaglanda of Ravenna and Amboglanna of the Notitia Dignitatum), but only the Catuvellauni and Atrebates adopted coinage. This suggests that the n2- people had not used coins when they lived on the Continent. Now, it appears to be generally accepted that coinage, in the form of potins and gold staters, came into common use in southern and eastern England in the second half of the 2nd century BC, so it follows that the above tribal groups who used the hill-letters r, l1 and n2 must have come to Britain earlier than 150BC. But of course what matters is not when the coins became common in southern and eastern Britain, but when they became common on the Continent, or at least those parts of the Continent which were the places of origin of the migrants who came to Britain, and that was presumably earlier. It therefore seems fair to conclude that the last of the tribes mentioned above, i.e. those who used the hill-letter n2, came to Britain quite a few years before 150BC. And of course those who used the hill-letters r and l1 came even earlier (and those who used the hill-letters n1, s and m still earlier). But it is possible that those who used the hill-letter l2 arrived later, perhaps even towards the end of the 2nd century BC. These people seem to have used only inversion-type names. It is likely that they came from a region of the Continent where coins were not widely used, since otherwise one could reasonably expect Iron-Age coins to be found in those parts of northern Britain settled by l2-people, e.g. along the south coast of the Firth of Forth or in the area north of the Humber (the Parisi). Some Iron-Age coins have indeed been found to the north of the Humber, but most of them have been attributed to the Coritani/Corieltauvi, who lived south of the Humber.
2.11 Having established in paragraphs 2.1 to 2.9.3 who lived where in southeast England in the 2nd and 1st centuries BC it is now time to move on to see how this knowledge helps us understand more clearly the origin and distribution of the various kinds of potin coinage.
3 Kentish Primary Series
3.1 The coins of this type are often referred to as Thurrock type coins because a hoard of some 2000 of them was found near Thurrock in Essex in 1987, but here they are referred to as Kentish Primary Series coins, following Holman (Holman 2017, 206). There are three important questions to ask. Who manufactured the coins? When did production start? When did production cease? There has long been contention as to whether the coins were manufactured in Britain or were imported from the Continent. In the interests of brevity the text below simply assumes that they were minted in Britain.
3.2 As to who produced the coins it seems quite clear that they were minted by the people who used the hill-letter m and lived in southeast Kent, which is the area in which the greatest concentration of these coins has been found. But another, if smaller, concentration of these coins is seen in northwest Kent (Haselgrove 1995, 119, Fig. 55), another area populated by m-people, as explained in paragraph 2.1. There were evidently links with other people using the hill-letter m and living in other parts of southeast and east England, since one coin of this kind was found during excavation of Maiden Castle in Dorset (Celtic Magno). Another was found near Chichester and hence near the Trundle (another Celtic Magno). Other finds were at South Ferriby on the Humber (Celtic Gambrand) and at Bridgham to the east of Thetford (which appears to have been Celtic Camborico, the predecessor of the Romano-British Camborico of Iter V of the Antonine Itinerary). And Haselgrove indicates that some Kentish Primary Series coins have been found along the course of the river Trent (Haselgrove 1995, 119, Fig. 55). One would expect there to have been m-people along the course of the Trent to account for the presence of the corresponding river-letter r in Trisantona, the old name of the river.
3.2.1 Regarding the many coins found at Thurrock it seems most likely that these had been brought across the Thames from Kent for safekeeping at the time of the arrival of the r-people from the Continent or possibly the arrival of the n2-people (the Canti). The hoard of 300-400 Kentish Primary Series coins said to have been found near Gravesend (Haselgrove 1995, 119), in Kent itself, may well have been buried around the same time.
3.3 As to the start of production it seems most probable that potin coins, like the gold coins, were first imported into Britain and were then copied here. Haselgrove refers to the coin found at Maiden Castle and indicates that it was found in a deposit underneath a layer containing a hearth last used in about 150BC (Haselgrove 1995, 119). There is apparently some doubt as to whether the coin really belongs to the deposit in which it was found but if it really does belong below that layer including a hearth last used around 150BC (as Haselgrove appears to believe) then evidently the coin was minted earlier than 150BC. This seems consistent with the archaeological evidence, which indicates that the ramparts of Maiden Castle were rebuilt on a larger scale around the middle of the 2nd century BC (see, for example, the entry for Maiden Castle on the Pastscape website). This rebuilding was presumably carried out by the n2-people, in this case the Atrebates, who took the hillfort over from the earlier m-people, the name of the fort then becoming Magno, an old-style compound in the hill-letters m and n2. It thus seems reasonable to conclude that the Kentish Primary Series potin coin found at Maiden Castle was used by the m-people who lived there earlier than 150BC. But if we accept that Kentish Primary Series coins were used by m-people (i.e. by people who used the hill-letter m), if we accept that Kentish Primary Series coins reached all the places mentioned above in paragraph 3.2 while those places were still controlled by m-people (i.e. before they were taken over by people who used the hill-letter r, l1 or n2) and if we accept that the people who used the hill-letters r, l1 and n2 did not use coins before they came to Britain (discussed above in paragraph 2.10), then it would appear that the advent of Kentish Primary Series coins has to be pushed back quite some way into the first half of the 2nd century BC. But then we need to find some of these coins in deposits which can be fairly accurately and reliably dated to that period, and we need metal analyses to be performed on any such coins found to make sure that they are made of metal which was manufactured in Britain. Then we could say with some confidence that the coins were actually minted in Britain, most probably in southeast Kent, and not just imported from the Continent.
3.4 As to the date of cessation of production of the Kentish Primary Series potin coins the best evidence is probably provided by the history of the manufacturing centre recently discovered at the top of St. Thomas’s Hill in the northern outskirts of Canterbury (Clark and Lane 2015, 1). This centre may have been the mint, or one of the mints, which produced the potin coins. It is believed that the settlement was probably founded in the first centuries of the first millennium BC but that activity on the site decreased dramatically from about 100BC onwards, the hilltop then being given over to agriculture or stock-raising (Clark and Lane 2015, 1). Presumably industrial activity came to an end when eastern and southeastern Kent were overrun by Trinovantes displaced from Essex by the eastward expansion of the Catuvellauni (as explained in paragraph 2.6 above). The Trinovantes took over a swathe of land right round the southeast of Kent as far as the river East Rother, some of the m-people being pushed westwards in the direction of Hastings (Ravenna’s Mutuantonis) and the upper reaches of the East Rother river (as also explained in paragraph 2.6).
3.4.1 The argument set out in paragraph 3.4 finds some corroboration in Holman’s observation that Gallo-Belgic A and B gold coins are scarce finds in east Kent and that the import of gold coins only began to increase with Gallo-Belgic C in the early 1st century BC (Holman 2017, 210). This appears to indicate that the m-people who had occupied southeast Kent did not import gold coins from the Continent, whereas the Trinovantes, who took over southeast Kent around 100BC, did. That the Trinovantes imported gold coins from the Continent is borne out by the excavations at Heybridge in Essex, where one of the coins found was a Gallo-Belgic B gold stater.
4 Flat Linear I
4.1 The potin coins of this kind were divided by Allen into types A to L (Allen 1971), where the L type coins are believed to date to the period of Caesar’s campaigns in Gaul and Britain. Again the same questions should be asked. Who made the coins? When did production start? When did production stop?
4.2 As to who manufactured the Flat Linear I coins there seems little doubt that they originated in the north of Kent, the earliest finds - type A coins - being concentrated around the Medway to the south of Rochester, though other coins were later found further east (Holman 2017, 208). The m-people had lost their independence, having been overrun by the Trinovantes displaced from Essex by the Catuvellauni. The Flat Linear I coins must therefore have been produced by those r-people who still lived in the area east of the Medway, or by the Trinovantes who had come to Kent from Essex. But it is unlikely that they were minted by the Trinovantes. It is true that Kentish Primary Series coins and Flat Linear I coins have been found in Essex, but, leaving aside the hoard found at Thurrock, not in large numbers. It may be, as suggested by Holman (Holman 2017, 231), that metal detector users have been less active in Essex than in Kent or that coins have been found in Essex but have not been so well recorded as in Kent, but as things stand one has the impression that potin coins played a very minor role in the coinage system of Essex around 100BC. Nor does it seem likely that Trinovantes would suddenly develop a liking for potin coins on crossing the Thames estuary to Kent. It seems quite clear, then, that the Flat Linear I potin coins were minted by the r-people of northeast Kent, east of the Medway. Some support for this argument is given by the fact that a concentration of Flat Linear I potins has been found in East Sussex, in an area including Mount Caburn hillfort, just east of Lewes. This is identified on this website as Ravenna’s Raxtomessa, given as a river-name in Ravenna but quite clearly the land-name Racstomessa in the hill-letters m, s and r. The most recent element of the Celtic name is the inversion-type element rac meaning ‘hill steep’. In other words this part of East Sussex appears to have been inhabited by r-people, just like the people of north Kent east of the Medway. And so far as the present writer is aware the only Flat Linear potin coins found in this area of East Sussex are Flat Linear I coins, seven of them at Mount Caburn itself (Haselgrove 1995, 124, Fig. 60; 126). Note that the Rac element of Racstomessa is inversion-type and so could well be contemporaneous with inversion-type l1-place-names (and the corresponding river-names) in southeastern Kent and with the inversion-type element Mut of Mutuantonis at Hastings. It is thus entirely possible that the r-people of East Sussex had come from the eastern part of northeast Kent when some of the Trinovantes migrated from Essex to Kent.
4.3 The start of production of the Flat Linear I coins was probably related to the cessation of production of the Kentish Primary Series coins, discussed above in paragraphs 3.4 and 3.4.1. The m-people had lost their independence and the Trinovantes apparently had no interest in maintaining the production of potin coins. But the coins were evidently of some use and value and so the r-people started production. The mint will not have been at St Thomas’s Hill since this was now abandoned and the area around it was presumably in the hands of the Trinovantes. The new mint may have been in Rochester, close to the find spots of a number of Flat Linear I type A coins.
4.4 Production of the Flat Linear I coins by the r-people no doubt ceased when these people lost their independence, apparently shortly after Caesar’s return to the Continent in 54BC. The r-people in northeast Kent had no doubt been weakened by Caesar’s military action in 55 and 54BC and so the opportunity presented itself for Cassivellaunus to bring the r-people under his control, as discussed in paragraph 2.9.2 above.
5 Flat Linear II
5.1 The Flat Linear II potin coins are the types M to P in Allen’s classification (Allen 1971). As a percentage of the total number of Iron Age coins found the Flat Linear II coins are most common in the Cantian part of Kent, the area west of the Darent, at 20%. In the area between the Darent and Medway they make up about 2% of total Iron Age coins found, in northeast Kent some 10.5% and in east Kent, the area east of the river Stour, including the Isle of Thanet, 5.7%. Admittedly the total number of Iron Age coins found west of the river Darent is very small, 59 up until 2017 (Holman 2017, 121), but nonetheless the differences between the percentages quoted above are striking. In addition, in recent years a number of Flat Linear II coins have been found north of the Thames, in areas normally associated with the Catuvellauni and, to some extent, with the Trinovantes.
5.2 As to who produced Flat Linear II coins it seems reasonably clear that it was the Catuvellauni who took over responsibility for minting the potin coins. They had become the dominant power in southeast England before Caesar’s arrival in 55BC and it was the Catuvellaunian king Cassivellaunus who led the joint British forces against Caesar. Cassivellaunus appears to have brought the r-people, including those who lived in the Canterbury area, under his control, as explained in paragraphs 2.7.1 and 2.9.2 above. He presumably from that moment on controlled the issue of the potin coins. It is conceivable that the mint which had produced the Flat Linear I coins, perhaps at Rochester, continued to operate under Catuvellaunian control, but the coins from this point on were smaller and thicker than the late Flat Linear I coins, this perhaps indicating that they were produced in a new mint. That mint may have been in the Cantian (= Catuvellaunian) area of northwest Kent, west of the river Darent, but it seems more likely that the new coins were produced north of the Thames, nearer to the new centre of Catuvellaunian power, the new settlement of Verlamion down the slope from Prae Wood, just west of modern St Albans.
5.3 As explained in paragraphs 2.7.1 and 2.9.2 the Catuvellauni appear to have taken control of the r-people in northeast Kent soon after Caesar returned to the Continent in 54BC, so it seems safe to assume that production of the Flat Linear II coins started soon after that event. Until new coins became available the inhabitants of the new Averno down on the low ground near the river Stour will have continued to use Flat Linear I type L coins. Indeed Haselgrove (writing in 1995) tells us that almost all of the 61 potins found at Canterbury are of the types L to P (Haselgrove 1995, 124).
5.4 As to when production of Flat Linear II coins ceased it is difficult to give a precise date. Haselgrove, writing in 1995, tells us that the Flat Linear II coins found in east Essex greatly outnumber the Flat Linear I coins (Haselgrove 1995, 124, Fig. 60) and the Flat Linear II coins are most likely to have been used in that area after Colchester fell to the Catuvellauni. But it is not clear when that happened, though it must have been at the latest during the reign of Tasciovanus since he minted coins in both Verlamion and Camuloduno. But we cannot be absolutely certain as to the opening and closing dates of the reign of Tasciovanus. Haselgrove suggests that production of Flat Linear II coins probably ceased before the end of the 1st century BC (Haselgrove 1995, 122), though his own histogram (Haselgrove 1995, 122, Fig. 58) might suggest that production continued at least until the early years of the 1st century AD.
6 Miscellaneous points
6.1 When did the changeover from old-style to inversion-type names occur?
Note that the Atrebates took over Maiden Castle in Dorset sometime around the middle of the 2nd century BC (as explained in paragraph 3.1.2) and at that time they were still using old-style names (the gn element of the name Magno of Maiden Castle). Note also that east and southeast Kent were overrun by Trinovantes, the industrial centre on St. Thomas’s Hill by Canterbury going into dramatic decline and being abandoned sometime around 100BC or a little later. By that time the Trinovantes had changed over to using inversion-type names, as had the r-people of northeast Kent who moved to East Sussex on being displaced by the Trinovantes, and as had the m-people who moved west to Hastings and the upper reaches of the East Rother river, again on being displaced by the Trinovantes. It thus seems safe to conclude that the changeover from old-style names to inversion-type names occurred during the second half of the 2nd century BC.
6.2 Amminus and his DUNO coins
Holman refers to late pre-Roman silver and bronze coins of Amminus bearing the mint mark DUNO, the main distribution of these coins being in the area east of Canterbury (Holman 2017, 216). This is a very strange distribution and calls for an explanation. It is possible that many of the then residents of the area east of Canterbury fled as Caesar approached with his army in 55 and 54BC, leaving the area substantially unoccupied. It is further possible that when Cassivellaunus seized control of the Canterbury area soon after Caesar’s departure he may have settled some Catuvellauni in the area east of Canterbury. The coins of Amminus would then easily be explained. The coins were minted in Colchester for use by the Catuvellauni living in the area east of Canterbury. It is true that earlier coins used the mint mark CAM or CAMUL for Colchester, but there can surely be no objection in principle to using the DUNO part of Camuloduno as a mint mark instead, especially as duno was the element which the Catuvellauni themselves had added to Trinovantian Camulo at Colchester.
6.3 Razing Celtic Averno to the ground to make way for Roman Duroaverno
Following on from paragraph 2.9.2 above it may be noted in passing that the new settlement founded by the Catuvellauni on the site of the later Romano-British Duroaverno may have been largely destroyed, or abandoned, during the Roman invasion of AD43, since it would appear that the Romans razed what was left of the settlement to the ground and built a new town in the Roman style on the same site from scratch. This new Roman town was of course the town we know as Duroaverno. The Duro element appears to have been used in southeast England in the sense of ‘new Roman town at or near’ and then what followed (in the present case averno) was the name of the Celtic predecessor of the Roman town. A hint at the razing of the settlement to the ground is provided in the report of the Canterbury Archaeological Trust on excavations at the Marlowe Arcade site in Canterbury (Clark and Macintosh 2015, 10), though admittedly the report is speaking only of one particular small area within the late Iron Age settlement.
6.4 The Iron Age coins of Canterbury
Since the Catuvellauni seized control of the Canterbury area after Caesar returned to the Continent one should expect to find Catuvellaunian coins, i.e. Flat Linear II coins, in the new settlement built on the low ground by the river Stour. And indeed Haselgrove tells us that nearly all of the 61 British potin coins found at Canterbury are of types L to P (Haselgrove 1995, 124). The coins of types M to P are of course Flat Linear II coins, whereas type L is the last type of Flat Linear I, i.e. the coins produced by the r-people of northeast Kent. It would seem then that the r-people living in Bigbury Camp used type L coins (which might indicate that they hadn’t been long at Bigbury Camp before Caesar’s invasion) and that they took their L type coins with them to the new Averno down by the river Stour.
6.5 The Canti
The name Canti originally applied only to the n2 -people of northwest Kent, but later, after the Canti/Catuvellauni had taken control of the r-people of northeast Kent and the Trinovantes of southeast Kent, it is possible that the term Canti was applied to all the inhabitants of Kent, with the possible exception of the l2 -people living between the Darent and the Medway. But it is possible that these l2 -people had been brought under the control of the Canti/Catuvellauni. It is not clear whether the name Cantium promontory used by Ptolemy, and identified by Rivet and Smith as the South Foreland in Kent, had been applied by the Canti/Catuvellauni themselves or was invented by the Romans.
[This page was last modified on 16 May 2021]