PREVIOUS WORKFootnote ¹

In 1864 John Evans argued that Cunobelinus' bronzes could be divided into three typological classes: those with Cunobelinus' name alone; those on which it appeared in conjunction with the name of his father Tasciovanus; and finally, those on which it occurred with the name of the place of mintage, Camulodunum.Footnote ² The next attempt at typological ordering was that of Derek Allen (later adopted by Mack), who divided Cunobelinus' bronzes into an early group, which contained coins supposedly similar to those of Tasciovanus and employed ‘crude, inartistic’ designs that were not copied from Roman sources; a middle group, displaying greater Roman influence; and a later group, which represented a decline and a reversion to non-Roman types.Footnote ³

The discovery of large numbers of Cunobelinus' bronzes during excavations at Harlow temple in the 1960s prompted Allen to attempt a re-analysis of the bronze series.Footnote ⁴ Allen now argued that they could be divided into three groups; the first was a broad group containing early coins of somewhat ‘primitive’ appearance, generally similar to the coins of Tasciovanus. Most of these bore abbreviated legends of Cunobelinus, while some carried a Camulodunum legend as well as, or instead of, the ruler, and some bore the names of Cunobelinus and Tasciovanus. The second and third groups were later than the first, though apparently contemporary with each other, and showed ‘marked Roman influence on subject and treatment’. Coins in the second group bore legends proclaiming Cunobelinus to be the son of Tasciovanus, whilst those in the third group included the mint name Camulodunum. Allen noted that coins with Camulodunum legends were distributed in and around Colchester, which he argued was their mint site, whereas coins combining Cunobelinus and Tasciovanus legends were found further west and were probably minted at St Albans. He made no statements concerning the distribution or likely mint site(s) of the early coins which lacked either Camulodunum or Tasciovanus legends. In the following two decades Allen's scheme was generally accepted, and indeed it received support, to a point, from Colin Haselgrove's chronological analysis of stratified finds in 1987.Footnote ⁵ Both Warwick Rodwell and Haselgrove suggested a slight alteration to the scheme: that the types with Cunobelinus' name alone were the earliest issues.

In 1989 Van Arsdell published Celtic Coinage of Britain in which he argued, in contrast to Allen, that there were no regional subdivisions in the coinage of the ‘North Thames’ region, within which he believed coins circulated freely.Footnote ⁶ He divided Cunobelinus' bronzes into an early group of types dated c. a.d. 10–20 and a late group dated c. a.d. 20–43. Van Arsdell's views on the homogeneous nature of coin distribution in the ‘North Thames’ region have been widely accepted in the literature during the following decades and this notion has also been supported by Haselgrove's and John Creighton's work, in which the ‘Eastern kingdom’ has been outlined as a major coin-issuing polity, although Haselgrove himself was aware of regional distinctions within Cunobelinus' bronze.Footnote ⁷

The works discussed above principally examined Cunobelinus' bronze in terms of typology, distribution, chronology, and political and tribal histories. In the 1970s and 1980s, however, archaeologists such as John Collis and Haselgrove began to argue that the focus of research should shift towards examining how Iron Age coins were used in Iron Age societies and how they were deposited in archaeological contexts.Footnote ⁸ To this end Haselgrove conducted an analysis of the archaeological contexts of Iron Age coins from South-East England and concluded that, in contrast to gold coins, struck bronzes were concentrated at the major nucleated settlements (with the key exception of Harlow temple), which he argued meant that they probably had a market function.Footnote ⁹ In 2000 Creighton also examined how British Iron Age rulers used coin iconography, in particular images derived from the classical world, to increasingly exert their power and authority over their subjects.Footnote ¹⁰

In this paper new light will be shed on the topics that — as seen in the discussion of previous work — are of the greatest significance to understanding Cunobelinus' bronzes. These are typology, chronology, distribution, political history, imagery, find context and function. Despite Collis' and Haselgrove's critiques, understanding the typology, chronology and distribution of Iron Age coinages is fundamental and must form the underlying basis of all other research upon them.Footnote ¹¹ An attempt will be made here to resolve the significant debate on the typological and chronological ordering of Cunobelinus' bronze types. In addition, their distribution, which has been subject to limited examination in the past, will be explored.Footnote ¹² Haselgrove grouped Cunobelinus' bronzes together with either all the other struck bronzes, or with all the other contemporary coins, when conducting his broader analyses on the archaeological contexts of Iron Age coins in South-East England. More specific analyses of their archaeological contexts and function will be conducted in this paper. Creighton's ideas about how Cunobelinus' coins were used to exert power and display identity will also be built on in the light of a more nuanced and detailed understanding of typology, chronology and distribution.

The ever increasing quantity of principally metal-detected coins recorded at the Celtic Coin Index (CCI) throughout the 1990s and 2000s, and in recent years by the Portable Antiquities Scheme (PAS) as well, means that the current work has a much greater pool of data to work from than previous studies. In addition, several important recent relevant works can be taken into account, such as Mark Curteis' and David Holman's regional surveys of Iron Age coinage in Northamptonshire and east Kent respectively and Philip de Jersey's work on Cunobelinus' silver.Footnote ¹³ This study also uses the new type catalogue for British Iron Age coins, Ancient British Coins (ABC), which although it does not discuss the ordering of Cunobelinus' bronzes in any detail, provides an excellent classification and outline of the other British Iron Age coin types.Footnote ¹⁴

TYPOLOGY

Cunobelinus' bronzes are divided here into five typological groups: an early and a later Eastern North Thames group, an early and a later Western North Thames group, and a Kentish group. It will be shown later on in this paper that these groupings are strongly supported by distributional analyses (indeed, the groups have been named on the basis of their distributions) and, to some extent, by analyses of metallurgy, metrology and stratigraphy. This study does not involve die linking, since the poor condition of most of the coins would render any attempt to do so extremely difficult and it is almost impossible to die link the vast majority of coins. No closer chronological ordering of the coins has been attempted beyond the broad groupings proposed. The findings suggest that the typological ordering of the bronzes proposed by Derek Allen is essentially correct, though a separate Kentish grouping has been established.

FIG. 1.

Early Eastern North Thames types.

LATER EASTERN NORTH THAMES TYPES

ABC 2945

• Obv. Griffin stepping r., holding snake in mouth, standing on exergual line. CA above and MV in front of griffin (on a single example the CAMV legend is replaced by an upside-down CVNOBELINVS legend above the griffin). Pellet border.

• Rev. Horse rearing r. Quatrefoil above, crescent above this. Dolphin above and to the l. of the horse. CVN below. Pellet border.

• No. of coins in CCI and PAS database: 62

ABC 2975

• Obv. CVNO in tablet, surrounded by wreath. Pellet border.

• Rev. Horse stepping r. CAMV below, on top of exergual line. Pellet border.

• No. of coins in CCI and PAS database: 98

ABC 2978

• Obv. Bearded head of ‘Jupiter/Zeus Ammon’ l. Head has horns. CVNO in front, BELIN behind. Pellet border.

• Rev. Warrior on horse r., wearing helmet, short sword in r. hand, circular shield in l. hand. CAM below, on top of exergual line. Pellet border.

• No. of coins in CCI and PAS database: 83

FIG. 2.

Later Eastern North Thames types.

ABC 2981

• Obv. ‘Janus’ heads, one facing l., the other r. CVNO in tablet below. Pellet border.

• Rev. Sow on haunches r. Tree behind. CAMV in tablet below. Pellet border.

• No. of coins in CCI and PAS database: 81

ABC 2984

• Obv. Bearded head of ‘Jupiter/Zeus Ammon’ r. Head has horns. CVNO in front, curved line behind. Pellet border.

• Rev. Lion crouching r. Tree behind lion. CAM in tablet below. Pellet border.

• No. of coins in CCI and PAS database: 151

ABC 2987

• Obv. Sphinx rearing r., teats visible. CVNO below, above exergual line. Pellet border.

• Rev. Man (‘Perseus’) standing l. Gorgon's head in r. hand, club in l. hand. CA to l., M to r. Altar behind ‘Perseus’. Pellet border.

• No. of coins in CCI and PAS database: 63

The later group of Eastern North Thames types is marked by a heavy emphasis on classical themes and designs and in particular by a distinctly classicising style, much closer to the styles and designs of the classical world than Cunobelinus' earlier coins are. Whilst precise classical prototypes for the designs cannot always be identified, it is clear that the die-cutters were fairly competent in classical engraving styles and had a thorough knowledge of classical imagery, though it should be borne in mind that classical imagery was not necessarily copied wholesale; as will be discussed later, it was probably carefully selected to suit Cunobelinus' political context and aims and the classical images employed may have been interpreted and understood quite differently in Britain than in the classical world.Footnote ³³ The classicising trend in Cunobelinus' later issues is also evident in his gold, where the latest ‘Classic’ type coins are very classical in style.Footnote ³⁴ It is also evident in the silver and, indeed, some of the designs on Cunobelinus' later Eastern North Thames silver types are closely paralleled on some of the later Eastern North Thames bronzes.Footnote ³⁵ Whilst de Jersey felt able to divide the later Eastern North Thames silver into two chronologically distinct groups, with the earlier featuring vegetal motifs and the latter employing figures and/or unusually small lettering in their inscriptions, there is no reliable basis on which to chronologically subdivide the later Eastern North Thames bronzes.Footnote ³⁶

A classical style is evident in the design of the griffin, and often in that of the horse, depicted on ABC 2945; furthermore, another classical motif, a dolphin, is also employed on the reverse of this coin, which as de Jersey has noted, closely resembles the dolphins on one of Cunobelinus' later silver units (ABC 2900).Footnote ³⁷ The wreath on the obverse of ABC 2975 is very similar to wreaths on several of Cunobelinus' later silver units (ABC 2876, 2891, 2894), and on a silver unit of Agr (ABC 3005), whose coinage is contemporary with the latter issues of Cunobelinus. ABC 2975 is almost identical to a bronze unit of Amminus (ABC 468), who issued coins in Kent during the latter part of Cunobelinus' reign.Footnote ³⁸

The heads on the obverses of ABC 2978 and 2984 are derived from classical images of the god Zeus/Jupiter Ammon. Possible prototypes include the depiction of Jupiter Ammon on an aureus and a denarius of Q. Cornuficius from the African mint, dated 42 b.c., and a range of denarii issued in Cyrenaica in 31–30 b.c., whilst Creighton has suggested that depictions on the coinage of the client king Juba I of Mauretania may also have served as a possible prototype.Footnote ³⁹ Whether any of these coins actually reached Britain during the Late Iron Age is debatable (at the time of writing there are none in the PAS database); but it is worth bearing in mind Creighton's argument that in the aftermath of Caesar's British campaigns, British aristocrats were sent to Rome (or other Roman cities) as hostages where they may well have come into contact with such coins. Of course, images of Jupiter Ammon may also have appeared on other media within the Roman world, including gems, which may possibly have reached Britain. De Jersey has noted the similarity of the horseman on the reverse of ABC 2978 to examples on one of Cunobelinus' later Eastern North Thames silver units (ABC 2888), and has suggested that some examples were almost certainly engraved by the same hand.Footnote ⁴⁰ The seated lion on the reverse of ABC 2984, meanwhile, is clearly a classical image, which may have been derived from a gem.Footnote ⁴¹ The tablet below the lion is similar to that below the boar on Cunobelinus' bronze unit ABC 2981, and on his silver unit ABC 2897, one of his later Eastern North Thames issues.Footnote ⁴²

The obverse of ABC 2981 is ultimately derived from depictions of the Roman god Janus, which occurred on numerous Roman Republican coin types. Van Arsdell has argued that the reverse of ABC 2981, depicting a tree behind a boar, was based on a bronze unit of Dubnovellaunos (ABC 345). The styles of these two coins are in fact completely different, though they may both have conveyed the same message/story to their Late Iron Age audience. The reverse of ABC 2987 has a scene derived from classical mythology (perhaps via a gem), that of Perseus and Medusa.Footnote ⁴³ Once again, this scene may have been interpreted quite differently by British viewers than by their classical counterparts.Footnote ⁴⁴

EARLY WESTERN NORTH THAMES TYPES

FIG. 3.

Early Western North Thames types.

ABC 2918

• Obv. CVNOBELINI in two panels (CVNOB or CVNOBE in upper panel, ELINI or LINI in the lower). Pellet in ring above and below. Pellet border.

• Rev. ‘Victory’ l., wearing helmet and seated on knobbed chair with back. Circular wreath with a pellet in the centre held in ‘Victory's’ r. hand. TASC below exergual line, though on many examples the latter part of this inscription curves up and behind Victory (some examples appear to read TASC F). Pellet border.

• No. of coins in CCI and PAS database: 141

ABC 2933

• Obv. Helmeted bust (with both shoulders visible) l., helmet with streamers. CVNOBELI in front. Pellet border.

• Rev. Boar on haunches l. Branch(?) in front. Trio of pellets above. TASC. FIL in between two lines below. Pellet border.

• No. of coins in CCI and PAS database: 45

All of Cunobelinus' Western North Thames coins (both early and late) can be distinguished from his Eastern North Thames and Kentish coins on the basis of their legends. In addition to bearing Cunobelinus' name, all of the Western North Thames coin types bear Tasciovanus' name (often abbreviated and frequently followed by ‘F’ or ‘Fil’ to convey that Cunobelinus was the son of Tasciovanus). In contrast, the name Tasciovanus is never present on the Eastern North Thames and Kentish coins, and this is the case with the silver too.Footnote ⁴⁵

The two coins in the early Western North Thames group can be stylistically separated from those in the later group because they are much less ‘classical’ in style than the later coins, though both of these coins utilise certain designs derived from the classical world. The obverse of ABC 2918 employs a double-panelled inscription, as does Cunobelinus' early Eastern North Thames bronze ABC 2924, though there are notable differences between the coins in terms of the inscription and decorative style. One of Cunobelinus' middle Eastern North Thames silver units (ABC 2858) employs a similar legend to ABC 2918, also arranged in double panels, though in this case stars are placed above and below the panels instead of pellets in rings; de Jersey has suggested that they were possibly inscribed by the same hand.Footnote ⁴⁶ The seated ‘Victory’ on the reverse of ABC 2918 is derived from Roman imagery, with a denarius of M. Porcius Cato, issued in Africa c. 47–46 b.c., proving a close match, including similar details such as the knobbed chair and the wreath held out in front, although Victory faces the opposite direction, both of her feet are on the ground and she does not wear a helmet.Footnote ⁴⁷ The seated ‘Victory’ on ABC 2918 is also similar to that on Cunobelinus' Eastern North Thames middle issue silver unit ABC 2855. However, the ‘Victory’ on the silver unit holds a bowl or vessel and de Jersey has argued that it is probably derived from different versions of M. Porcius Cato's denarii, or from denarii of M. Cato dated to 89 b.c., where Victory holds a patera rather than a wreath.Footnote ⁴⁸ The bust on the obverse of ABC 2933 is also based on a classical image, the plumed helmet possibly suggesting that it was derived from a depiction of Mars, or ‘Virtus’ (or similar), which occur on numerous Augustan and (in the case of the former) Republican coins, though the bust on ABC 2933 is engraved in a non-classical style.

LATER WESTERN NORTH THAMES TYPES

ABC 2957

• Obv. Laureate head l. CVNOBEL in front, I below, NI behind. Pellet border.

• Rev. Centaur r., blowing horn, standing on horizontal line. TAS below, CIOVA in front, NI. F above. Pellet border.

• No. of coins in CCI and PAS database: 199

ABC 2960

• Obv. Helmeted head r. CVNO behind, BELINVS in front. Pellet border.

• Rev. Sow r., pellet triad below head. TASCIIOVANII above. Boar stands on two horizontal lines with a line of pellets in between them. F below the lines. Pellet border.

• No. of coins in CCI and PAS database: 171

FIG. 4.

Later Western North Thames types.

ABC 2963

• Obv. Warrior on horse r., wearing helmet, holding spear in r. hand and oval shield in l. hand. CV behind, NOB below. Pellet border.

• Rev. Warrior standing l., wearing crested helmet, holding spear in r. hand and shield in l. hand. TASCIIO to left, VANTIS to right. Pellet border.

• No. of coins in CCI and PAS database: 222

ABC 2966

• Obv. Bust r. CVNOBE behind, LINVS REX in front. Pellet border.

• Rev. Bull r., with head lowered and front leg bent. TASC below exergual line. Pellet border.

• No. of coins in CCI and PAS database: 199

ABC 2969

• Obv. Winged head l. CVNO in front, BELIN behind. Pellet border.

• Rev. Seated ‘metalworker’ r. Vessel in l. hand, tool (hammer?) in r. hand. Figure set upon a horizontal line. TASCIO behind. Pellet border.

• No. of coins in CCI and PAS database: 330

ABC 2972

• Obv. Winged horse r. CV behind, NO below. Pellet border.

• Rev. ‘Victory’ stabbing bull r., with knife in l. hand. TASCI below. Pellet border.

• No. of coins in CCI and PAS database: 289

Cunobelinus' later Western North Thames bronzes are generally classicising in their design and style. Many of the designs employed on these coins are derived from the classical world, though there are also several designs that do not appear to have clear classical origins and are likely to have been ‘local’ in concept. The designs are all produced in a fairly classical style, though there are some ‘crude’ elements, particularly in the styles of ABC 2963, 2966 and 2972. Chronological subdivision within this group is difficult at present, though the ‘cruder’ coins may possibly be the earliest. There are some stylistic links to the Western North Thames silver issues.Footnote ⁴⁹

The laurel head on the obverse of ABC 2957 is derived from a portrait of Augustus, Tiberius, Caligula or Claudius (either of the first two are clearly more likely for chronological reasons). As Creighton has noted, the portraits of these four emperors are schematic and hard to differentiate from each other.Footnote ⁵⁰ The centaur on the reverse has no evident parallels with Republican or early Imperial coinage; indeed, no lone centaurs are depicted on the latter. Scheers suggests a coin of Magnetes, Thessaly, c. 196–146 b.c., as a prototype, though the centaur here is not blowing a horn.Footnote ⁵¹ A non-coin prototype, perhaps a gem, was probably employed. The portrait on the obverse of ABC 2960 is in classical style and wears a helmet in the style of the Imperial period. It has no parallels on Roman coins and, in fact, may not actually have been based on a classical prototype at all. A Roman helmet similar to the one depicted on this coin was found buried with the ‘Conquest period’ coin hoards at Hallaton in Leicestershire, which hints that Roman helmets may well have been present in Britain towards the end of the Late Iron Age.Footnote ⁵² Creighton has suggested that the sow on the reverse of ABC 2960 fits into the suite of imagery employed by Augustus to assert his authority, in this case the sow refers to the ‘Golden Age’, though a link to classical meanings is by no means necessary or certain.Footnote ⁵³ The mounted warrior depicted on the obverse of ABC 2963 may be derived from depictions of mounted warriors on coins of Tasciovanus (cf. ABC 2562, 2565, 2568, 2571, 2577, 2580, 2610, 2640 and, in particular, 2676) and Rues (ABC 2754). It is also typologically close to the horseman on a gold stater of Epaticcus (ABC 1343), who, like Cunobelinus, claimed to be Tasciovanus' son and issued coins in the Southern (South Thames) region, most likely during the latter part of Cunobelinus' reign.Footnote ⁵⁴ The posture and position of the horse and the placement of the legend on the obverse of ABC 2963 is extremely close to that of the winged horse (‘Pegasus’) and the legend on the obverse of ABC 2972. The soldier on the reverse of ABC 2963 is not entirely dissimilar to depictions of Mars (and perhaps Hercules) on certain Late Republican and Augustan coin issues and, as Creighton has noted, this image may well have been based on classical representations of Mars.Footnote ⁵⁵

The butting bull on ABC 2966 is similar to the bull on aureii and denarii of Augustus issued at Lyon 15–10 b.c.,Footnote ⁵⁶ whilst the head on the obverse is in a very similar design to the head on one of Cunobelinus' Western North Thames silver units (ABC 2882). On ABC 2969 the winged head on the obverse is close to a depiction of Medusa on a denarius of L. Cossutius from 74 b.c., and is perhaps even more closely paralleled by images on numerous classical gemstones.Footnote ⁵⁷ The ‘metalworker’ on the reverse of the same coin has similarities to examples on Kentish silver units of Dubnovellaunos (ABC 324, 333), though the depiction on Cunobelinus' coin is far more classical in style and execution and the vessel being worked upon is not present on Dubnovellaunos' coins.Footnote ⁵⁸ This scene is not obviously based upon classical prototypes; hammering workmen do appear on a group of second- and first-century b.c. classical gemstones, though there are no examples depicting a vessel being hammered.Footnote ⁵⁹ As Allen and Williams have noted, the metalworker on ABC 2969 may well have been a representation of a locally important smith god, or similar, though of course the true meaning of this scene eludes us.Footnote ⁶⁰ The unusual vertical representation of TASCIO on the reverse of ABC 2969 is closely matched on the reverse of one of Cunobelinus' Western North Thames silver units (ABC 2870).Footnote ⁶¹ The ‘Victory’ sacrificing a bull on the reverse of ABC 2972 is an image derived from the classical world, with an aureus of Augustus minted at Pergamum in 19 b.c. providing a close parallel.Footnote ⁶²

KENTISH TYPES

FIG. 5.

Kentish types.

ABC 2921

• Obv. Winged horse l., head turned back to right. Usually with CAM, or occasionally a pellet in circular wreath below (some examples may have neither). Pellet border.

• Rev. ‘Victory’ standing l., palm leaf in l. hand pointing downwards, uncertain object (possibly a snake) in r. hand. CVN in front. Occasionally pellet in ring behind. Pellet border.

• No. of coins in CCI and PAS database: 133

ABC 2930

• Obv. ‘Victory’ standing r., holding circular wreath in r. hand. C to l. V to r. Pellet border.

• Rev. Horseman holding reins in r. hand. Left arm often invisible. Pellet or another object above horse's head. CVN below. Pellet border.

• No. of coins in CCI and PAS database: 28

ABC 2936

• Obv. Horse r., head turned back, straps around neck and ribs. Pentagram below, star behind, pellet above, CVN in front. Pellet border.

• Rev. Horseman r., r. hand holding reins. Left arm often invisible. Pellets either side of man's head. CVN below and in front, each letter separated by one of the horses legs. Pellet border.

• No. of coins in CCI and PAS database: 6

ABC 2939

• Obv. Ship with sails and rigging. CVN below. Pellet border.

• Rev. ‘Victory’ walking r., holding wreath. S to l. E to r. Pellet border.

• No. of coins in CCI and PAS database: 20

A characteristic feature of Cunobelinus' Kentish bronzes is the use of a winged figure derived from the classical Victory, though it is not certain that ‘Victory’ was employed or interpreted in the same way by both British and classical audiences.Footnote ⁶³ ‘Victory’ is depicted on three of the four Kentish types (ABC 2921, 2930, 2939). As already seen, ‘Victory’ also appears on Cunobelinus' early Eastern North Thames bronze ABC 2927 and on his early Western North Thames bronze ABC 2918. It is also present on Kentish coins of Eppillus (ABC 387, 414) and Touto (ABC 432), which probably date to the early first century a.d., as well as appearing on Southern region coins of Tincomarus (ABC 1130), Verica (ABC 1214, 1241, 1244, 1247), and Epaticcus (ABC 1349). The depictions of ‘Victory’ on ABC 2930 and 2939 are particularly similar. On both of these coins ‘Victory’ wears a characteristic skirt or tunic, which is paralleled on the two coins of Eppillus.

Cunobelinus' Kentish bronzes have very strong stylistic links to other Kentish coinages, particularly those of Dubnovellaunos. The turned-back head, displayed by the winged horse on ABC 2921, is a popular feature amongst animals on Kentish coins. The closest parallels for the horseman on the reverse of ABC 2930 are the recently discovered gold stater of Anarevito found near Dover and the Tasciovanus silver unit (ABC 2625).Footnote ⁶⁴ The horse on ABC 2930 also looks like the Linear style horses of Cunobelinus' second gold series.Footnote ⁶⁵ The horse on the obverse of ABC 2936 has a similar posture and style to various animals on Kentish coins of Dubnovellaunos (ABC 333, 342 and 354), Sam (ABC 369 and 381) and Eppillus (ABC 423), one of Dubnovellaunos' Eastern North Thames bronzes (ABC 2404), and one of Cunobelinus' early Eastern North Thames bronzes (ABC 2915). However, the horse on ABC 2936 differs from all of these examples because its front legs are standing, rather than rearing or leaping. The obverse also features a pentagram, a symbol largely restricted to Kentish coins, for example uninscribed bronze units (ABC 282, 285) and Kentish coins of Dubnovellaunos (ABC 309, 354), though it also features on silver units of the Southern region rulers Tincomarus (ABC 1103) and Verica (ABC 1256) and on an uninscribed silver unit in Kentish style but apparently distributed in the Eastern North Thames region (ABC 2258). The horseman on the reverse of ABC 2936 also has an image from a Kentish coin as its closest match; in this case the horsewoman on the reverse of a bronze unit of Dubnovellaunos (ABC 348). The S and E on the reverse of ABC 2939 may possibly relate to Sego, which is featured on several Kentish coins, probably minted during the early years of the first century a.d., often combined with a Tasciovanus legend (ABC 441–53). Sego is probably a ruler, but may be a Kentish place name.Footnote ⁶⁶

DISTRIBUTION

The CCI and PAS databases combined hold records for 1,944 provenanced examples of Cunobelinus' bronze coins from 338 sites.Footnote ⁶⁷ They have an extensive distribution across a large part of South-East England ( fig. 6). The vast majority of examples are encountered in the region between the valleys of the rivers Thames and Nene and in Kent. They are much less common in areas that are dominated by coins belonging to different broad regional traditions, such as northern East Anglia (Icenian/East Anglian coinage), Hampshire, Sussex, Berkshire and Surrey (‘Southern’ coinage), the Cotswold region (Dobunnic/Western coinage), and the East Midlands (coins of the Corieltavi or North-Eastern coinage).Footnote ⁶⁸

FIG. 6.

Overall distribution of Cunobelinus' bronzes.

One notable gap in the distribution, which is not filled by coins of a different regional tradition, is Greater London. To some extent this probably relates to a lack of metal-detecting in recent decades in what is largely an urban area. However, a recent distribution map of ‘Celtic art’ in England and Wales and a map showing the overall distribution of Iron Age coins in England and Wales both suggest that some Iron Age metalwork finds are known from Greater London, and, indeed, Greater London does not appear to represent an obvious gap on either of these maps.Footnote ⁶⁹ A closer analysis of the Iron Age coins from London recorded in the CCI and on the PAS database reveals that these coins are virtually all cast potins, with some Gallo-Belgic gold and a few ‘Armorican’ imports. These coins are all fairly early, dating from the mid-second to mid-first century b.c. Footnote ⁷⁰ There are actually very few Iron Age coins post-dating the mid-first century b.c. This is an interesting finding and may well reflect a genuinely low level of activity in the London region during the latest part of the Iron Age. If correct, the reasons for this are at present unclear, though one might speculate that it may perhaps relate to the post-Caesarian political situation in the region.Footnote ⁷¹ Another gap in the distribution of Cunobelinus' bronze coinage is encountered in mid-Kent and in the Weald, though, apparently unlike London, this gap is also evident in the overall distribution of Iron Age coins and in the distribution of Celtic art. Indeed, Garrow notes that the Weald appears to have been sparsely populated and relatively ‘blank’ archaeologically throughout much of prehistory.Footnote ⁷²

Although the overall distribution of Cunobelinus' bronzes covers a large part of South-East England, an analysis of findspots of the individual coin types makes it clear that there are actually three distinct regional groupings, which coincide with the typological groupings outlined above. Whilst there are some overlaps in the distribution of the coins in each of these three groups, they are in fact remarkably clear and distinct from one another. The first group of types has a distribution focused on the Eastern North Thames region, which covers most of Essex and southern Suffolk ( fig. 7). The second group is found across the Western North Thames region, which is Hertfordshire, Bedfordshire, Buckinghamshire, Northamptonshire, southern and eastern Oxfordshire, parts of Cambridgeshire and the Essex–Hertfordshire border area ( fig. 8). The third group has a distribution focused on Kent ( fig. 9).Footnote ⁷³ The exact same pattern of regional subdivision is also encountered with Cunobelinus' silver.Footnote ⁷⁴ In the gold, meanwhile — aside from the earliest Biga issues, which were restricted to the Eastern North Thames region — all of the types circulated widely throughout Cunobelinus' territories and, judging from their legends, may all have been minted at Colchester.Footnote ⁷⁵

FIG. 7.

Distribution of Cunobelinus' Eastern North Thames bronzes (ABC 2912, 2915, 2924, 2927, 2942, 2945, 2948, 2951, 2954, 2975, 2978, 2981, 2984, 2987, 2990).

FIG. 8.

Distribution of Cunobelinus' Western North Thames bronzes (ABC 2918, 2933, 2957, 2960, 2963, 2966, 2969, 2972).

FIG. 9.

Distribution of Cunobelinus' Kentish bronzes (ABC 2921, 2930, 2936, 2939).

So how can the three strongly regional distributions that are observable in Cunobelinus' bronze (and silver) types be interpreted? Traditional ‘tribal’ models of interpretation, whereby regional coin distributions are equated with one of the British civitates as set out in Ptolemy's mid-second-century a.d. Geography, have been critiqued in recent years. The use of the ‘tribe’ as a social model for understanding Late Iron Age societies has been attacked as a product of colonial-era thinking and the back projection of Roman-period social groupings onto the Late Iron Age has also been critiqued.Footnote ⁷⁶ Instead, work by scholars, such as Haselgrove, Creighton, Mattingly and Leins suggests that ‘political’ interpretations are far more likely, with coin distributions in the Late Iron Age of South-East England relating not to tribes, but to the political power of individuals and political dynasties.Footnote ⁷⁷ On the basis of Haselgrove's work, Creighton has interpreted the overall principal distribution area of Cunobelinus' coinage as the political entity ruled by him, which he has termed the ‘Eastern Kingdom’.Footnote ⁷⁸

The existence of regional sub-groupings within the broader spread of Cunobelinus' coinage is perhaps curious. It is highly suggestive that the ‘Eastern Kingdom’ was in fact divided into three political sub-groupings, which endured throughout Cunobelinus' reign. As Creighton has argued, it is probable that political power in South-East England during the Late Iron Age was dynastic and that political power and ‘kingdoms’ could be inherited or passed from one ruler to another.Footnote ⁷⁹ Here it is argued that this provides the origin of the political sub-groupings evident in Cunobelinus' bronze and silver distributions.

The Western North Thames distribution corresponds extremely closely to the principal spread of Tasciovanus' coinage, whilst that in the Eastern North Thames region appears to correspond to the principal distribution of Dubnovellaunos' North Thames coinage.Footnote ⁸⁰ It is thus very likely that the two political sub-groupings evident in the distribution of Cunobelinus' bronze and silver north of the Thames were based on the political spheres of power of earlier rulers. How far back these ‘dynasties’ go is difficult to know in our present state of knowledge, though detailed study of the early British coinage may throw light on this issue.Footnote ⁸¹ Creighton has suggested that the North Thames dynasties might be related to the kings mentioned by Caesar — Cassivellaunus in the Western North Thames region and Mandubracius (and his father), of the Trinobantes, in the Eastern North Thames region — and to their struggle for power, which may have been further impacted by a hypothesised carving up of political power organised by Caesar in Britain following his invasions of 55 and 54 b.c. Footnote ⁸² Something along these lines may well be close to the truth.

The Kentish situation is somewhat more complex. Here inscribed coins appear to have been issued by a significant number of different rulers in a fairly short period before Cunobelinus' coins appear in this region.Footnote ⁸³ Clearly the political situation here was fluid and complex and, indeed, Caesar mentions that there were four kings in Kent.Footnote ⁸⁴ On the basis of the coinage, it can be suspected that Cunobelinus' accession to power in this region may have represented something new: a single king ruling in Kent.

At this stage it is pertinent to note that, despite the recent theoretical arguments to the contrary,Footnote ⁸⁵ the three regional distributions evident in Cunobelinus' bronze and silver types do appear to correspond, as far as the limited historical descriptions go, to three of Ptolemy's British civitates. The Eastern North Thames distribution could correspond to the Trinovantes, the Western North Thames to the Catuvellauni, and the Kentish to the Cantiaci.Footnote ⁸⁶ There does not seem to be any obvious reason why the Roman civitates in these three regions could not have been based upon the key political groupings present in these regions at the time of the Roman conquest. Whether the civitates names can be back-projected is perhaps more debatable; aside from Ptolemy, the Catuvellauni are only mentioned by Cassius Dio (writing in the early third century) in his description of the Claudian invasion and also on an inscription from Hadrian's Wall.Footnote ⁸⁷ Aside from Ptolemy, the Cantiaci are only mentioned in the Ravenna Cosmography and on a Roman votive tablet, though the term ‘Cantium’ was used as a geographical descriptor by Diodorus Siculus, Caesar and Strabo, all writing well before the Claudian invasion.Footnote ⁸⁸ The Trinovantes (Trinobantes), however, were mentioned by Caesar, which is suggestive that this group name endured throughout the Late Iron Age and into the Roman period.Footnote ⁸⁹

Whilst the regional coin distributions, discussed above, have been interpreted as relating to political groupings and individual and dynastic power, it should be borne in mind that there may well have been some sort of ‘ethnic’ dimension, or aspect of ethnic identity, associated with being part of the political élite in one or more of these regions. Perhaps even those who were subject to the rule of particular individuals or dynasties may have had some sense of ethnic identity associated with this. Mattingly has suggested that the term ‘peoples’ is more appropriate than the term ‘tribes’ when describing the identity of groups in pre-Roman and Roman-period Britain.Footnote ⁹⁰ Clearly studying Late Iron Age and Conquest-era social organisation and group identity is a fruitful topic that needs further examination in the future.

METALLURGY, METROLOGY AND MINTS

There have been several studies in which the metal compositions of small numbers of Cunobelinus' bronzes have been analysed. The full results are too lengthy to reproduce in this paper, but a simple summary is given in Table 1. As Peter Northover has noted, in general there appears to be little correlation between numismatic detail and alloy type; indeed, the variation in the composition of the alloys used for Cunobelinus' bronzes suggests that they may simply have been produced by melting down whatever copper-alloy objects were to hand, rather than by precisely controlling their composition, as was done with the precious metal coinage.Footnote ⁹¹ A pattern observable in Clogg and Haselgrove's results is for copper and low tin bronzes to dominate Cunobelinus' later Eastern North Thames bronzes, whilst mid- and, to a lesser extent, high tin bronzes are the dominant alloy used for the later Western North Thames bronzes. This finding might provide support for the suggestion that these two groups of coins were produced in different places. However, this must be tempered by the fact that Northover's analyses had low tin bronzes dominating the later Western North Thames bronzes.

TABLE 1.

METALLURGICAL ANALYSES OF CUNOBELINUS' ‘BRONZE’ COINS

(C = Copper, less than 1% tin; LTB = low tin bronze, 1–4% tin; MTB = mid-tin bronze, 4.1–9% tin; HTB = high tin bronze, more than 9% tin; BG = bronze/gunmetal, with 18% tin and 7.9% zinc)

Another potentially interesting finding is that the single analysed specimen of the early Western North Thames type ABC 2933 is a bronze/gunmetal, which distinguishes it from all the rest of Cunobelinus' ‘bronze’ issues, since it is the only one that contains notable amounts of zinc. The inclusion of zinc, ultimately derived from Roman brass, may link this type to the copper/brass and brass issues of Tasciovanus (ABC 2688, 2691, 2712), Andoco (ABC 2727) and Dias (ABC 2748, 2751), though ABC 2933 is notably different from these owing to its higher tin content.Footnote ⁹²

There is a considerable problem in assessing the weights of Cunobelinus' bronze coin types for, unlike precious metal coins, they suffer from significant wear and corrosion over time, which means there is a wide scatter of weights across a considerable range. There are simply not enough recorded weights to make any reliable assessment of most of the early Eastern North Thames and the Kentish issues, though the majority of types in these groups have their most common weights between 2.10 and 2.49 g.Footnote ⁹³ Exceptions to this rule include ABC 2912 and 2915, which appear to be a little light, with only 4 out of the 36 weighed examples of ABC 2912 and 3 out of the 25 weighed examples of ABC 2915 exceeding 2 g, with maximum weights of 2.22 and 2.13 g respectively. The only other bronze types with comparable weights are some of Dubnovellaunos' North Thames series (ABC 2404, 2407, 2413),Footnote ⁹⁴ which may perhaps support the suggestion (made above on the basis of typology) that ABC 2912 and 2915 were amongst Cunobelinus' earliest bronze issues. ABC 2990 and 2993 appear to be half units;Footnote ⁹⁵ ABC 2990 has a maximum weight of 1.17 g (10 examples), while ABC 2993 has a maximum weight of 1.27 g (3 examples). ABC 2942 may also be light since the highest of the 4 recorded weights is 1.45 g.

Of the early Western North Thames types ABC 2933 seems to be somewhat light, with the majority of examples weighing between 1.30 and 1.99 g, though 7 out of 42 examples have weights above 2 g, with a maximum weight of 2.54 g. The apparently low weight may perhaps link this coin with the earlier copper/brass and brass issues of Tasciovanus, Andoco and Dias, which also seem to have a pattern of generally low weights.Footnote ⁹⁶ ABC 2918 is notably different from the later Western North Thames types in that its most common weight category is 2.50–2.59 g, whilst it also has a small but notable proportion of coins with weights above 3 g.

All of the later Eastern North Thames types have a similar weight distribution, which allows them to be tentatively grouped together ( fig. 10). The most common weights are between 2.10 and 2.39 g, with the highest number in the 2.10–2.19 g decile. The later Western North Thames types can be similarly grouped together ( fig. 11) and they are clearly slightly heavier, with the most common weights falling between 2.20 and 2.49 g, with the highest number between 2.40 and 2.49 g. The weights of both the later Eastern and Western North Thames bronzes were probably quite closely controlled, though perhaps not with the same degree of precision as the precious metal coinages. Assessment of this issue is difficult owing to the aforementioned problems of wear and corrosion.

FIG. 10.

Weights of Cunobelinus' later Eastern North Thames bronzes (ABC 2945, 2975, 2978, 2981, 2984, 2987).

FIG. 11.

Weights of Cunobelinus' later Western North Thames bronzes (ABC 2957, 2960, 2963, 2966, 2969, 2972).

It is highly likely that most, perhaps all, of Cunobelinus' Eastern North Thames bronzes were minted at Colchester (Camulodunum),Footnote ⁹⁷ though production at other as yet unidentified mints should not be ruled out. Camulodunum is recorded on the legends of many of the Eastern North Thames coin types and almost all of these types have been found at the site in fair numbers. It is also the only site in the Eastern North Thames region where mint debris has been discovered. Some of the numerous slab moulds identified at Colchester were found to have been used for the production of bronze coins.Footnote ⁹⁸

As has been seen, the typological, distributional and, to some extent, the metallurgical and metrological evidence all show that the Western North Thames types are noticeably different from the Eastern North Thames issues and hence were probably produced at locations within the Western North Thames region. The two best mint candidates are Braughing/Puckeridge and St Albans (Verulamium), though bronzes may possibly have been produced at other unidentified mints as well. Mint debris has been recovered from both sites. Analysis of some of the 40 kg of coin moulds from Braughing/Puckeridge suggests that they were used to make a tin bronze coinage with a small fraction of silver, which would fit with the majority of Cunobelinus' Western North Thames bronzes, though it also corresponds with some issues of Tasciovanus.Footnote ⁹⁹ There appears to have been a decline in occupation at Braughing/Puckeridge from c. a.d. 25, though considerable numbers of Cunobelinus' later bronzes are known from the site (71 later Western North Thames and 7 later Eastern North Thames units out of a total of 97 examples of Cunobelinus' bronze).Footnote ¹⁰⁰ Analysis of some of the coin moulds recovered from Late Iron Age contexts at St Albans demonstrates that bronze coins were also produced at this site.Footnote ¹⁰¹ Of course some of the mint activity at both of these sites could relate to earlier rulers, such as Tasciovanus, though it can be suspected that one, or both, of them, functioned as a mint during Cunobelinus' reign.

Turning to the Kentish issues, it can be argued that ABC 2921 was minted at Colchester since it bears a Camulodunum legend and, although the majority of examples are found in Kent, there are also a notable number of coins of this type from Colchester and other sites in Essex and Suffolk. The other Kentish issues (ABC 2930, 2936, 2939) do not bear a Camulodunum legend and their distributions are virtually restricted to Kent. On this basis, it can be suggested that they may have been minted somewhere in Kent.

POLITICAL HISTORY, IMAGERY AND POWER

Creighton has argued that Iron Age coin typologies and distributions have often been used to create political histories of Iron Age rulers that cannot be sustained by the limited evidence and should be dismissed as conjecture and myths.Footnote ¹⁰² It is argued here, however, that it is possible to use the broad knowledge of the typology and distribution of Cunobelinus' coinage to come to some cautious conclusions about his political history. In addition, it is proposed that this knowledge of typology and distribution can be utilised to examine how Cunobelinus used coin imagery to reinforce and legitimate his power at different stages during his reign and in the different regions under his control.

On the basis of his coinage, it is suggested that Cunobelinus' reign can be divided into three broad relative phases, though it is considered unlikely that there were sharp cut-offs between these phases and some coin types may have straddled two different phases. The first phase is clearly evidenced in the gold; Cunobelinus' first gold issues (the Biga types) are confined to the Eastern North Thames region, which stands in sharp contrast to his four later gold issues, which have a distribution extending to the Western North Thames region and into Kent.Footnote ¹⁰³ De Jersey has also argued that Cunobelinus' earliest silver types were also restricted to the Eastern North Thames region.Footnote ¹⁰⁴ This is suggestive of a short gap between Cunobelinus' initial accession to power in the Eastern North Thames region and the extension of his rule to the Western North Thames region and Kent.Footnote ¹⁰⁵ How long this phase lasted is difficult to estimate, but it was evidently short and something in the order of a few years is the most likely possibility.Footnote ¹⁰⁶ It may be suspected that Cunobelinus began minting bronze at this time as well and that some of his early Eastern North Thames bronzes belong to this early stage. However, as discussed, it is difficult to disentangle the precise chronologies of the early Eastern North Thames bronze issues and it is not possible to divide those that might have belonged to the first stage of Cunobelinus' rule from those that belonged to the second stage, though, for reasons discussed above, there are hints that ABC 2912 and 2915 may well have been particularly early issues.

Cunobelinus very probably succeeded Dubnovellaunos (who also issued apparently contemporary coins in Kent) as ruler in the Eastern North Thames region.Footnote ¹⁰⁷ The precise nature of his accession is unknowable. It may have been through inheritance or peaceful means, but bearing in mind that Dubnovellaunos appears to be recorded (as Dumnobellaunos) in Augustus' Res Gestae as a British king who (at an unspecified date) suppliantly sought refuge with Augustus, it may be suspected that Dubnovellaunos was ejected from his throne.Footnote ¹⁰⁸ The range of styles employed on Cunobelinus' early Eastern North Thames coins does not shed much light on this problem. The styles employed on these coins include a mixture of influences, with links to Kent (including Kentish coins of Dubnovellaunos) and to the Western North Thames region.Footnote ¹⁰⁹ Given that Cunobelinus claims to be the son of Tasciovanus on his Western North Thames issues, it is possible that he conquered the Eastern North Thames region, ousting Dubnovellaunos, with the support of his father Tasciovanus, but this cannot be proven for certain and further speculation would be unwise.

The coin legends are, however, particularly instructive. Abbreviated legends of Camulodunum feature on a notable proportion of the bronze and silver coins issued by Cunobelinus in the Eastern North Thames region throughout his reign.Footnote ¹¹⁰ This tells us that control of Camulodunum was probably central to Cunobelinus' identity and status in the Eastern North Thames region from the very beginning of his reign onwards. Camulodunum appears to have been an important site for several decades prior to Cunobelinus' accession; the archaeological evidence suggests that significant occupation began at the site during the mid- to late first century b.c.,Footnote ¹¹¹ whilst two of Tasciovanus' gold issues (ABC 2559 and 2589) bear legends that appear to read ‘Camvl’. These coins are amongst Tasciovanus' earliest issues, though the findspot evidence for these coins, and for Tasciovanus' coinage in general, suggests that Tasciovanus probably did not control or mint coins at the site of Camulodunum itself.Footnote ¹¹²

Cassius Dio, referring to the events of the Claudian invasion, describes Claudius' capture of Cunobelinus' (Cynobellinus') capital, Camulodunum, as the culmination of the initial phase of the Roman invasion.Footnote ¹¹³ This is certainly suggestive that Camulodunum was a very important site at the time of the Roman invasion, though caution is required in taking the Roman sources at face value, since its importance may have been exaggerated to make Claudius' brief involvement in the invasion of Britain appear more significant than it actually was. Camulodunum legends also appear on all of Cunobelinus' gold issues, which following the initial stage of his reign circulated beyond the confines of the Eastern North Thames region into the Western North Thames region and Kent. This tells us that whilst Camulodunum was of great symbolic importance in the Eastern North Thames region, it also became a symbol of authority that had some wider significance in all of Cunobelinus' domains.

During the second phase of his rule, Cunobelinus increased his power and domains enormously, extending his rule to the Western North Thames region and to Kent and, in doing so, coming to rule the major part of South-East England. This expansion is evidenced in Cunobelinus' gold, since starting with his second (Linear) series his gold is no longer restricted to the Eastern North Thames region, but its distribution encompasses the Western North Thames region and Kent as well.

The date of Tasciovanus' reign in the Western North Thames region has been estimated by several authorities at c. 20 b.c.–a.d. 10, though of course this is not exact.Footnote ¹¹⁴ It is highly likely that Cunobelinus succeeded Tasciovanus as ruler of the Western North Thames region. Cunobelinus included an abbreviated name of Tasciovanus on all the bronze and silver coins he issued in the Western North Thames region throughout his reign. On many coins he records himself as the son of Tasciovanus. This tells us that portraying a relationship to the previous ruler Tasciovanus was central to Cunobelinus' image in the Western North Thames region. Cunobelinus legitimised and reinforced his right to rule through reference to Tasciovanus and claimed to be his son.Footnote ¹¹⁵

Cunobelinus issued four bronze and two silver types for Kentish use. Both de Jersey and Holman have suggested that Cunobelinus' Kentish issues may have been minted into the latter part of Cunobelinus' reign.Footnote ¹¹⁶ This can be queried, for the styles of the Kentish coins are clearly early and if they were minted for a long period of time, one might have suspected that classicising styles, as became current on Cunobelinus' later Eastern and Western North Thames types, might have been introduced. The number of Kentish types issued is also fairly small. It is more likely that Cunobelinus' Kentish issues were restricted to a fairly early period. Following this, he probably handed over control, for a time at least, to his son Amminus who issued small quantities of coins in Kent during the latter part of Cunobelinus' reign.Footnote ¹¹⁷ A characteristic feature of Cunobelinus' Kentish bronzes is the use of Victory, which is depicted on three of the four types (ABC 2921, 2930, 2939). It can be suspected that this imagery may relate to Cunobelinus' victorious takeover in Kent, perhaps through military force.Footnote ¹¹⁸ However, it should be borne in mind that Victory may have had a different meaning in a British as opposed to a Roman context. The prevalence of Victory on Kentish coins, not just on those of Cunobelinus, may well suggest that it represented some sort of locally important deity.

The third phase of Cunobelinus' reign saw him continue to rule in the Eastern and Western North Thames regions, though, as discussed, he may have handed over power in Kent, for a time at least, to his son Amminus. The classical sources tell us that he was dead at the time of the Claudian invasion in a.d. 43, though he appears to have still been alive at the time of Caligula's aborted invasion of Britain in early a.d. 40.Footnote ¹¹⁹ Cunobelinus' later ‘Plastic’ and ‘Classic’ gold issues are distributed across the Eastern and Western North Thames regions and Kent, whilst he issued silver and, particularly, bronze coins in the Eastern and Western North Thames regions in vast quantities. The iconography of his later silver and bronze increasingly incorporated classical styles and imagery. As shown for the bronze, in many cases the classical prototypes appear to have been fairly old, dating to the Late Republican or Augustan period. Whilst it is possible that there was a sudden influx of classical prototypes, such as gems, towards the latter part of Cunobelinus' reign, earlier Iron Age coin imagery suggests that similar classical prototypes had been known for some time previously. It is perhaps more likely that there was a change in the way that Cunobelinus chose to express his power and status. He may have chosen increasingly to adopt classical styles and images of power.Footnote ¹²⁰ Of course, another explanation could be that there was an influx of classically-trained die-cutters to Cunobelinus' kingdoms towards the end of his reign.

ANALYSIS OF FINDSPOTS

The sites on which Cunobelinus' bronzes have been found are divided into a number of broad site type categories, which will be outlined and discussed in detail below ( figs 12–14). This has been done on the basis of the archaeological evidence available for the find sites, but it is important to note that there are certain problems in doing this. Firstly, whilst some coins have been found during excavations,Footnote ¹²¹ a large proportion of the coins are metal-detected finds and (with the notable exception of coins recorded via the PAS) finders have often typically only provided the CCI with fairly vague provenances, usually giving the nearest town or village to the findspot. This can make assignation of coins to precise archaeological sites a somewhat difficult matter. Furthermore, defining discrete site types is also a problematic issue, since certain sites changed in usage over time and could fit into more than one category; for example some oppida/central places, such as Colchester and Saham Toney, also had a Roman military presence, whilst oppida/central places, such as St Albans and Colchester, also appear to have had temples situated within their wider complexes. These limitations should be kept in mind during the following discussion.

FIG. 12.

Number of Cunobelinus' bronzes at different site types (total provenanced coins = 1,944).

FIG. 13.

Site types yielding Cunobelinus' bronze. 1. Colchester (Essex); 2. Heybridge (Essex); 3. Silchester (Hants.); 4. Baldock (Herts.); 5. Braughing (Herts.); 6. St Albans (Herts.); 7. Canterbury (Kent); 8. Saham Toney (Norfolk); 9. Thetford (Norfolk); 10. Abingdon (Oxon.); 11. Dorchester-on-Thames (Oxon.); 12. Chichester (W Sussex); 13. Harlow (Essex); 14. Springhead (Kent); 15. Worth (Kent); 16. Rothwell Top (Lincs.); 17. Farley Heath (Surrey); 18. Longthorpe (Cambs.); 19. Richborough (Kent); 20. Old Winteringham (Lincs.); 21. Alchester (Oxon.); 22. Wroxeter (Shrops.).

FIG. 14.

Distribution of ‘productive sites’ with five or more examples of Cunobelinus' bronze. 20 coins: 1. Evenley (Northants.). 10–19 coins: 2. Sandy (Beds.); 3. Chelmsford (Essex); 4. Fingringhoe (Essex); 5. Great Chesterford (Essex); 6. Great Dunmow (Essex); 7. Ashwell (Herts.); 8. Cow Roast (Herts.); 9. Broadstairs (Kent); 10. Goodnestone (Kent); 11. Shorne (Kent); 12. Little Waldingfield (Suffolk). 5–9 coins: 13. Odell (Beds.); 14. Haversham-cum-Little Linford (Bucks.); 15. Newport Pagnell (Bucks.); 16. West Wycombe (Bucks.); 17. Worminghall (Bucks.); 18. Kelvedon (Essex); 19. Langenhoe (Essex); 20. Orsett (Essex); 21. Saffron Walden (Essex); 22. Little Hadham (Herts.); 23. Therfield (Herts.); 24. Acol (Kent); 25. Cliffe (Kent); 26. Eccles (Kent); 27. Minster, Thanet (Kent); 28. Ramsgate (Kent); 29. Waldershare Park, Dover (Kent); 30. Duston (Northants.); 31. Little Milton (Oxon.).

The first category considered is oppida/central places. These sites are highly variable and do not conform to a consistent type.Footnote ¹²² They are archaeologically evidenced by various combinations of dykes, high status burials, continental imports, industrial activity, coin flan moulds and concentrations of Iron Age coins.Footnote ¹²³ It is probable that they will all have had a significant central function for surrounding settlements, serving variously as political, market, religious and population centres. The following is a list of Late Iron Age oppida/central places with Cunobelinus' bronzes, together with the number of his bronzes that can be assigned to each site: Colchester = 205; Braughing = 97; Baldock = 44; St Albans = 42; Heybridge = 28; Dorchester-on-Thames = 14; Canterbury = 9; Silchester = 7; Abingdon = 5; Chichester = 4; Saham Toney = 4; Thetford = 1.

Large numbers of Cunobelinus' bronzes derive from oppida/central places, with 23.7 per cent of the total coming from these few sites. Within this category the largest number of coins comes from Colchester and also three major sites in Hertfordshire: Braughing, St Albans and Baldock, sites that have all been the subject of large-scale excavations in which the majority of the coins known from these sites were found.Footnote ¹²⁴ There is a notable group of coins from Heybridge, which is situated on the Essex coast and has considerable excavated evidence for continental imports, probably playing an important role as an entrepôt for continental contact and trade.Footnote ¹²⁵ Smaller numbers of Cunobelinus' bronzes have been found through excavation and metal-detecting at (or near) Abingdon and Dorchester-on-Thames (Dyke Hills), both of which have substantial dykes enclosing areas of occupation, but which have not yet been excavated on the scale of the aforementioned sites.Footnote ¹²⁶ Canterbury, meanwhile, is the best known oppidum/central place in Kent, though only a relatively small number of Cunobelinus' bronzes have been found there.Footnote ¹²⁷ This is perhaps surprising, but no reason for this can be advanced in the present paper. Small quantities of bronzes have also been found at oppida/central places outside Cunobelinus' territory, such as Silchester and Chichester, which were two of the major sites of the Southern dynasty.Footnote ¹²⁸ A handful of coins are also known from a couple of the oppida/central places in Icenian territory, Saham Toney and Thetford.Footnote ¹²⁹

The second site type category is temples. These are sites of a principally religious nature where coins were deposited for votive reasons. The following is a list of temple sites with Cunobelinus' bronzes, together with the number of his bronzes found at each site: Harlow = 744; Worth = 28; Springhead = 9; Rothwell Top = 1; Farley Heath = 1.

40.3 per cent of Cunobelinus' bronzes derive from temples, though most of these in fact come from excavations at a single site, Harlow. Harlow is located on a hilltop, apparently near the boundary of the distribution of Cunobelinus' Eastern and Western North Thames bronzes, which may have been a significant factor in its location, though Western North Thames bronzes are much more common than Eastern North Thames bronzes (cf. figs 7–8). The votive nature of the site is evident from the large numbers of Iron Age coins, brooches, bracelets, finger-rings and pins deposited there, and a temple was constructed during the Flavian period. Although a few uninscribed gold coins were probably deposited there in around the mid-first century b.c., the overwhelming majority of coins were deposited during a short phase in the latter part of Cunobelinus' reign, or, more likely, after the Roman conquest. This is supported by the coin types, which are dominated by Cunobelinus' later bronzes, and by the wear patterns on the coins, with earlier coins showing evidence of much greater wear. Most of the coins were found in the Roman temple layers and were evidently disturbed from earlier deposits, but many coins were found in pre-temple deposits in a layer interspersed with mid-first-century a.d. brooches and animal remains. A large group of coins was found in association with a circular gully, which belonged to a circular structure. This gully had silted up prior to the deposition of the coins, which suggests some sort of votive activity related to an old building.Footnote ¹³⁰

Of the other temple sites, metal-detecting at Worth has uncovered over 200 Iron Age coins and hundreds of Roman coins. The site occupied a promontory surrounded by marshland and it appears to have functioned as a votive site between the mid-first century b.c. and the Late Roman period and a temple was constructed here in the fourth century a.d. Footnote ¹³¹ Springhead was situated at springs marking the source of the river Ebbfleet and it functioned as a major religious complex and focus for religious depositions throughout the Roman period, whilst over 80 Iron Age coins have also been found at the site.Footnote ¹³² Metal-detecting at Rothwell Top has revealed a votive site characterised by large numbers of Iron Age and Roman coins as well as votive miniatures.Footnote ¹³³ Farley Heath was a Roman temple situated in flat country in Surrey and several Iron Age coins have been found here.Footnote ¹³⁴

A handful of Cunobelinus' bronzes have been found at Roman forts, though these represent only a very small proportion of the total (0.9 per cent). The relevant sites are: Old Winteringham = 6; Alchester = 6; Richborough = 3; Longthorpe = 2; Wroxeter = 1.

The remaining coins have not been assigned to precise site types as such, since to do so would require a major study in its own right. Whilst a small number of coins in this remaining group have been found in excavations at Late Iron Age or Romano-British settlements that cannot as yet be classified as Late Iron Age oppida/central places the vast bulk are metal-detected finds.Footnote ¹³⁵ As already discussed, there are considerable problems with working out exactly what sites most of these metal-detected coins came from, whilst most of the finds reported under the PAS would require an in-depth study in order to ascertain the nature of the sites from whence they derived. The author suspects that the majority of the remaining coins did come from settlement sites of some form, or from sites associated with industrial activity, though some may have come from temples/religious sites as well. In order to settle this issue more surveys are clearly needed along the lines of Holman's and Curteis' regional studies in which the nature of coin-producing sites are tackled in greater detail.Footnote ¹³⁶

The remaining coins have, however, been divided into two groups: ‘productive sites’ and ‘other’. Productive sites are locations (or localities) from which five or more examples of Cunobelinus' bronze have been recorded, and which are probably worthy of further detailed study. 273 coins, representing 14 per cent of the total, derive from productive sites ( fig. 14), whilst 410 coins, representing 21.1 per cent of the total, come from the ‘other’ sites, which have fewer than five examples of Cunobelinus' bronzes.

Finally, it is important to note that, in contrast to Late Iron Age precious metal coinages, there is an apparent lack of hoarding of Cunobelinus' bronzes. There are currently no known ‘landscape hoards’ of Cunobelinus' bronzes, that is hoards placed in the landscape away from the oppida/central places and temples.Footnote ¹³⁷ This is an interesting phenomenon. Perhaps bronzes were considered to be too ‘low value’ to be hoarded in this way. If hoarding in the landscape represents religious deposition of precious metal coinage, then a similar motivation may lie behind the bronzes found at temples and at religious sites within the oppida/central places.

FUNCTION

The function of Cunobelinus' bronzes will now be examined, firstly by attempting to explain why they were issued in the first place. One possibility is that Cunobelinus' bronzes were issued to help facilitate market exchange. Strabo, writing in the early first century a.d., implies the existence of market relations between Britain and the Continent when he provides a list of goods imported to and exported from Britain and notes that heavy tax duties were paid on these goods, though he makes no direct mention of coinage with respect to this commerce.Footnote ¹³⁸ The author believes that the large quantities of products imported from the Roman world into South-East Britain from the late first century b.c. onwards arrived largely through trade, though it is possible that some arrived as a result of diplomatic journeys made between South-East Britain and the near Continent.Footnote ¹³⁹

Cunobelinus' bronze coinage was, however, almost certainly not involved in the direct purchase of goods imported from the Continent. His bronzes did not make their way to the Continent in any quantity and the small numbers that did are probably best accounted for as the personal possessions of individuals who had crossed the Channel. Instead, if the bronzes were indeed issued to facilitate market exchange then it would have been for regional market exchange within Cunobelinus' domains. The concentration of bronzes at oppida/central places is certainly not inconsistent with the idea that the bronzes were used in market exchange, since these sites would have been well suited to function as market centres and, indeed, the imported continental goods are concentrated heavily at these sites.Footnote ¹⁴⁰

Issuing bronzes to facilitate market exchange may have brought a further benefit to Cunobelinus and his regime. It seems probable that much Roman bronze was put into circulation by the state through money-changers, who exchanged bronzes (useful for low value transactions) with individuals for precious metal coins that had initially been put into circulation by the state making payments to its soldiers and officials.Footnote ¹⁴¹ The state could then get precious metal coins back from the money-changers. Issuing bronze in this way could, therefore, have been an effective mechanism for Cunobelinus and his officials to get back some of the precious metal coin issues ready for reminting.

The significance of market exchange in Britain prior to the Roman conquest has, however, recently been questioned by some scholars.Footnote ¹⁴² Haselgrove and David Wigg-Wolf have also argued that Iron Age coinage was never used as general purpose money and that Iron Age coins were largely bound up in various kinds of ceremonial or religious usage and never broke out from the confines of traditional societies and their value systems.Footnote ¹⁴³

Indeed, it is perhaps more likely that, as with the precious metal coinages, Cunobelinus' bronzes were issued for social and political reasons.Footnote ¹⁴⁴ Cunobelinus and his officials could have arranged for bronzes to be struck in order to pay or reward certain followers and supporters. Issuing large quantities of bronze beside precious metal coins would potentially have enabled Cunobelinus and his officials to make a wide range of low-level payments, perhaps to different groups of people, or for different reasons or minor services. It is possible that they would have been handed out at important ceremonies held at the major central places to reinforce Cunobelinus' authority and to create bonds between himself and his people. A potential parallel for this proposed social function comes from northern Gaul, where during the second half of the first century b.c. bronze was struck in massive quantities.Footnote ¹⁴⁵ In northern Gaul bronzes did not serve to facilitate Roman-native trade, for only one late issue is found in Roman Augustan forts in quantity. Instead they may have replaced some of the functions of the earlier native gold and silver coinages, which had largely disappeared following their removal by the Roman state during and after the Gallic Wars. Of course, in considering this it should be borne in mind that, unlike the tribes of northern Gaul, Cunobelinus was able to issue precious metal coins in large quantities.

It has been shown that large numbers of Cunobelinus' bronzes were votive deposits at temple sites and Curteis has argued that Iron Age coins were almost always deposited in ritual contexts, whether at temples, oppida/central places or other settlements.Footnote ¹⁴⁶ There is indeed evidence that some of Cunobelinus' bronzes from oppida/central places were ritually deposited; for example, four of the coins from Colchester were excavated from Gosbecks temple enclosure ditch, whilst the coins from Saham Toney were concentrated in a stream near a bridge.Footnote ¹⁴⁷ It is highly unlikely, however, that the bronzes were principally issued in order to play a role in religious deposition. The vast numbers of coins deposited at Harlow temple may all have been deposited in the post-conquest phase.Footnote ¹⁴⁸ An examination of Cunobelinus' bronzes from dated contexts (excluding those from Harlow) also indicates that the vast majority were deposited after the Roman conquest, though it should be borne in mind that some coins in later deposits may have been disturbed from earlier occupation ( fig. 15). This is the case for both Cunobelinus' early and later issues, though a much greater proportion of the early coins were deposited in pre-Roman contexts, a fact that lends further support to the typological divisions outlined above.

FIG. 15.

Date of the contexts in which Cunobelinus' early (above) and later (below) bronzes were deposited (excluding Harlow temple). (Based on data from Haselgrove (Reference Haselgrove1987, 359–478) with additions. Sites used from Haselgrove (Reference Haselgrove1987): Odell (Beds.), Colchester (Essex), Kelvedon (Essex), Beauchamps, Wickford (Essex), Silchester (Hants.), Baldock (Herts.), Braughing (Herts.), St Albans (Herts.), Canterbury (Kent), Chichester (W Sussex), Houghton Regis, Puddlehill (Beds.), Abingdon (Oxon.), Lower Nazeing, Nazeingbury (Essex), Eccles (Kent), Dorchester-on-Thames (Oxon.). Additional data from: Wavendon Gate (Bucks.) (Williams et al. Reference Williams, Hart and Williams1996), Silchester (Hants.) (Fulford and Timby Reference Fulford and Timby2000), Gravelly Guy (Oxon.) (Lambrick and Allen Reference Lambrick and Allen2004), Bourne Hill, Wherstead (Suffolk) (Gill et al. Reference Gill, Plouviez, Symonds and Tester2002).)

It is important to note the strong regional tendencies in the distribution of Cunobelinus' bronze types, which it has been suggested reflect three political sub-groups under Cunobelinus' control. Whilst there is certainly overlap between the coins issued for the different groups (particularly the Eastern North Thames and Kentish coins), the general regional distinctiveness of the distributions is striking and significant. This suggests that there were strong social and political factors governing the issue and subsequent circulation and exchange of the bronzes.

Finally, it is important to stress that, in addition to their exchange functions, bronzes were also issued for another important secondary reason, namely to allow Cunobelinus to use imagery to assert his authority, power and identity, which, as has been discussed in detail above, was tailored to suit specific regional and temporal contexts.Footnote ¹⁴⁹ Issuing ‘low value’ bronzes beside precious metal coins meant that this imagery could potentially reach a wider portion of society than precious metal coins alone could. In addition, a tripartite coinage, comprising gold, silver and bronze issues would in itself have served as symbol of status; by issuing a large tri-metallic coinage, Cunobelinus, like Tasciovanus before him, was very probably consciously following the precedent of the Roman Imperial coinage, which was officially issued in gold, silver and bronze from c. 23 b.c. onwards.Footnote ¹⁵⁰