Roman coins from Bornholm - a preliminary overview more

Helle W. Horsnæs Roman coins from Bornholm – a preliminary overview Fig. 1. Number of coin finds (total amount and Roman coins) recorded in Denmark from 1840 till 2009. Bornholm is extremely rich in archaeological finds, and when it comes to Roman coins Bornholm presents the undisputed highest find density in all Denmark1. Today ca. 2500 ancient coins have been registered from this relatively small island, equal to 4.25 coins per km2, while the remaining parts of Denmark can boast only ca. 2750 coins, equalling 0.06 coins per km2. 1 The impact of Detector Age archaeology The richness of finds is partly due to the many skilled amateur archaeologists from Bornholm, who during the last quarter of a century have conducted numerous surveys with their metal detectors. Due to the intensive use of detector archaeology the number of coins found in Danish soil has risen dramatically. The number of Roman coins known has more than doubled within the last 25 years (Fig. 1). e absolute numbers of coins are given as rounded approximate figures. Since the beginning of the 1980s an average of ca. 150 Roman coins found in Denmark is registered in e Royal Collection of Coins and Medals of the National Museum in Copenhagen each year. 433 HELLE W. HORSNÆS It is, however, not merely a matter of more finds. The find pattern itself is gradually changing. Detector archaeology has yielded an entirely new set of sites, namely the “Black Soil” sites with huge numbers of artefacts found in the plough soil. The “Black Soil” areas are characterized by the very dark soil created by human occupation and manure from the livestock of the pre- or protohistoric settlement. The areas are also easily identified by chemical analyses revealing high phosphate contents – again due to the manure of the livestock. These sites, evidence of extensive pre- or protohistoric settlements, were hardly known 25 years ago. Detector archaeology has given us a huge number of single finds of coins and/or coins from scattered hoards. Roman denarii are among the common finds from the detector sites, and in these cases the coins have often been found spread over a large area and mixed with many other find types. The number of Roman gold coins known from Bornholm until 1980 was 128 – collected during 140 years i.e. little less than one solidus found per year, while the last quarter of the century has yielded 67 coins, ca. 2.5 solidi per year. The figures are of course not statistically valid, but they still give the clear indication that the average number of gold coins found per year has more than doubled. The “old” finds are mainly hoards of several coins found together, whereas the new finds are divided between hoards and single finds. The denarii present a much more eloquent picture. 402 denarii were found in the period 1840-1980 (2.9 denarii per year), again, often in the form of hoards that had been preserved almost intact until they were found. The period 1980-2006, however, has yielded no less than 1691 denarii (65 denarii per year). Thus the increase in the average number of denarii found per year has multiplied by more than 20 and by far exceeds the increase in the number of finds of solidi (Fig. 2). The fact that the number of silver coins found has risen considerably more than the number of gold coins is probably due to visibility. Finding a gold coin with a naked eye is far easier than finding a silver coin, and many gold finds have been made by a farmer working his field. Since World War II the increasingly heavy machinery employed in modern farming has taken the farmer away from direct contact with the soil, and at the same time the plough goes deeper into the topsoil thereby destroying cultural layers that had otherwise been intact. The layer of plough soil has become thicker, and the artefacts ploughed up are now being retrieved Fig. 2. Number of solidi and denarii finds from Bornholm in the years 1840-2006. 434 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW by the metal detector. And unlike the farmer, the metal detector “sees” silver as well as gold. The huge difference in the relative increase in the number of gold vs silver coins should warn us that silver coins, as well as other types of silver objects, may be under-represented in the finds from areas, where the metal detector is not habitually used in archaeology. All the coins have been registered in The Royal Collection of Coins and Medals with information on the date of issue of the coin, i.e. the name of the reigning emperor, and the majority of the finds have been identified with a type number according to the type catalogue Roman Imperial Coinage (RIC). This is a basic registration, which allows only for some preliminary conclusions. The present article draws on the information available in The Royal Collection of Coins and Medals. The information is gathered in a dynamic database, which is being constantly updated. The last entries to be considered here were made in May 2007. It is hoped that the database will eventually be accessible to the public on the Internet. An in-depth analysis of the coin finds in their contexts is much needed. The following comments draw on a limited number of randomly chosen case studies2. This is a good, operational and valid way to distinguish between find types. A mere count of coins would have led to a distorted picture if any statistical evaluation of the material was attempted, as one large hoard would inevitably blur the picture otherwise represented by a number of single finds. Moreover, coins found as single finds are often of a different character than hoarded coins. The interpretation of the function of a hoard is in itself a complicated issue. In the case of Roman coins in Bornholm, as well as in other parts of Barbaricum, hoards are most likely to be considered as deliberate deposits, either as saving hoards of precious metal, to be used in the future in a barter trade, or as raw material, or as a sacrifice, not to be retrieved by humans3. Single finds are often interpreted as the result of accidentally lost coins which can be seen as evidence on everyday use of the coins in, for example, trade. Contrary to the coins found in hoards, single finds of coins often tend to be of small denominations. However, in most parts of Barbaricum small denominations are found extremely rarely, and this is also the case in Bornholm (cf Fig. 3)4. The division between single finds and hoards is easy to undertake, when a hoard is found in situ – as the ideal “pot of money”. Detector archaeology, however, has provided us with many coins found one by one over a long period of time, but in the same site. In some cases, as for example the large Smørenge Hoard, an amateur archaeologist surveying with a metal detector initially found a small number of denarii, but few other finds. When more coins came up the local museum was called in to undertake an excavation that eventually revealed the presence of a ploughed up hoard (Kromann & Watt 1984). In this case 3 Hoards and single finds – a major methodological problem Traditionally, numismatists divide coin finds into two main groups: single finds and hoards. Coins found until ca. 1980 are published on find lists in Breitenstein 1944 and Kromann 1983-1984. Since then a limited number of coins have been published in various articles and/or in AUD (Arkæologiske Udgravninger i Danmark – Archaeological Excavations in Denmark, published 1984-2002). References are made here to the inventory number of the coin in e Royal Collection of Coins and Medals in National Museum in Copenhagen (FP), the registration number of the site in the Bornholms Museum (BMR no.), and the 6-digit parish number and site number (sb) in the Fund og fortidsminder database: http://www.dkconline.dk/ 2 e numismatist would also include the possibility that a hoard may be accidentally lost, as e.g. money in a purse, lost but not recovered by the original owner. 4 H. Horsnæs (2006a) discusses the finds of Roman small nominations in Denmark. 435 HELLE W. HORSNÆS Fig. 3. Nominals represented in coin finds from Bornholm. the original find-spot of the hoard was located, and it was revealed that originally the denarii had been deposited in two small pots. Thus there is no doubt that many of the Smørenge denarii were part of the hoard, but were all the coins part of the hoard? And what about a Late Roman solidus found close to the pot? In the years following the excavation of the Smørenge Hoard amateur archaeologists continued to find Roman denarii in the same field, and the distribution map of the detector finds seemed to indicate another cluster of coins in the same field. In 2000 a second excavation was undertaken, but although more coins were found within the same area, it was not possible to identify the find-spot (Vennersdorf 2002). Were the coins from the second cluster of finds originally a small part of the Smørenge Hoard that has been moved away from the centre of the hoard area in a lump of the clayey soil of the area? Or do they belong to another hoard, “Smørenge II”, the original find-spot of which, however, cannot be located? We will probably never find a definite answer to this question. Amateur archaeologists have also surveyed fields next to the field in which the Smørenge Hoard(s) was discovered. In these cases we are dealing most probably with single finds of denarii, as well as of other coin types from more recent periods. It is therefore possible that a number of coins from the hoard area were not part of the Smørenge Hoard, although the majority (but not all!) of the coins found in this area are denarii of the same types as the ones from the hoard(s). Smørenge is only one example of an important methodological and interpretative problem that we have to deal with when working with detector finds. The problem is relevant for all areas where intensive detector surveying is undertaken, but the great masses of finds from the Black Soil sites on Bornholm and the meticulous registration of them, stresses the importance of dealing with this issue. A viable procedure may be a combination of numismatic analysis (coins types, striking periods, mints, secondary treatment), analysis of distribution patterns, and an analysis of the surface wear of individual coins5. Still, we will probably have to accept that clear-cut definitions may not be reached, and that therefore we must get used to working with approximate numbers rather than precise absolute numbers of coins from hoards/single finds. Nominal types As hinted above, two coin types dominate the finds from Bornholm. The most commonly found Roman coin is a denarius from the 2nd c. AD, the second most common type is the Late Roman solidus from the 5th and the earliest part of the 6th c. AD (Fig. 3). 5 A case study is presented in Horsnæs 2006b. 436 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW Fig. 4. Chronological distribution of denarii found on Bornholm. Denarii Today more than 2200 denarii have been found on Bornholm. The overall chronological distribution of the denarii is similar to the pattern noted in other parts of Barbaricum: the material is dominated by denarii struck during the reigns of the Antonine emperors (Antoninus Pius, Marcus Aurelius and Commodus), denarii struck during the reigns of Trajan and Hadrian are quite common, while denarii struck during the Flavian and Severan dynasties are regularly found. Denarii from the early 1st c. (Julio-Claudian dynasty) as well as early 3rd c. denarii are rare, but not unseen (Fig. 4). Among these coins I am sure that the majority eventually will turn out to be interpreted as parts of (more or less!) well-defined hoards: 217 denarii Udmarken6 7 Robbedale 255 denarii 8 Smørengegård 1983 497 denarii Smørengegård 1984-2006 212 denarii found in the same field as the Smørengegård 1983 Hoard. A considerable number may belong to the hoard(s)9. 16+ denarii Borresø10 Sorte Muld11 281(?) denarii 12 Sylten 4 103(?) denarii Both Sorte Muld and Sylten 4 are sites related to the Ibsker parish central place13. In this parish a total of 698 denarii have been found in more than 30 “sites”, the majority of which belong to the central place. In this case the word “site” implies a registration unit, as e.g. a single field. Other sites in this cluster have yielded considerable numbers of denarii, and there are several other possible hoards in this material, e.g. Sylten 214 and Dalshøj 115. There are great differences in the preservation of the denarii. Quite a large number of the denarii from Bornholm are heavily worn (e.g. coins from the Smørenge Hoard), but this does 6 7 Breitenstein 1944, 3-7. Site 060301 sb 66. Breitenstein 1944, 7-33. FP 636. Site 060304 sb 20. 8 Kromann & Watt 1984. FP 4218. Site 060305 sb 144 9 Cf Vennersdorf 2002; Horsnæs 2006b. Breitenstein 1944, 33-35. e National Museum in Copenhagen, Dept. of Danish Prehistory inv.no. 2567. Site 060305 sb 154. See also Mørkholm 1969. 11 281 denarii have been registered from BMR no. 1191. Also other types of coins, including coins of somewhat later periods, have been registered as BMR 1191. Site 060403 sb 93. 12 BMR 1077. Site 060403 sb 168. e denarii probably derive from a scattered hoard. 13 On the Sorte Muld central place see now Adamsen et al. 2009. 14 55 denarii have been registered from BMR 789. Site 060403 sb 74. 15 53 denarii have been registered from BMR 1639. Site 060403 sb 135. 10 437 HELLE W. HORSNÆS a b Fig. 5a-b. Denarius with a hole first pierced later closed with a tap of silver: a – obverse, b – reverse. not apply to all the finds. Coins have corroded in various ways that are due most probably to variations in the chemical composition of the soil in which they were deposited (Horsnæs 2006b). There are a number of cases where the coins seem to have been burnt. This for example is the case of the denarii from Sylten 4 mentioned above16. In a few cases the denarii had been reworked in antiquity. At least 15 denarii were pierced, and often the piercing is placed in a rather illogical position. These denarii were clearly not intended to be used as pendants showing the face of the emperor. A pierced denarius was recently found in the Sorte Muld field in the Ibsker central place. The quite unusual piercing was made from the reverse of the coin with an instrument with a square section. The piercing was situated in front of the standing personification on the reverse of the coin, and behind the neck of Diva Faustina17. In one case a denarius struck by Commodus in AD 183 had been pierced, but later the hole was closed with a small tap of silver (cf Fig. 5)18. 16 A quite uncommon case of a denarius mounted in a silver frame has been found at Brogård19. The coin was struck by Otho in AD 69, and is thus one of the earliest denarii found in Bornholm. A coin with a similar mounting – but made in bronze – has been found in a C1b grave in Skovgårde in Zealand (Ethelberg 2000, 89). Solidi No less than 208 Late Roman solidi were found in ca. 40 sites on Bornholm20. Among these all except one were struck in the period after the division of the Roman Empire in AD 395. The only exception is a solidus struck at Milan in 379-388 AD, during the reign of Theodosius I (the Great). The latest issues represented in the material from Bornholm are solidi struck during the reign of the first Byzantine emperor, Anastasius (491518 AD)21. 19 BMR 1077. Site 060403 sb 198. Horsnæs 2007. e coin is a hybrid type: Diva Faustina Major/COS III. BMR 1191x2071R. FP 7422.1. 18 Site 060401 sb 160. FP 7254.3. BMR 3185x 38. 17 AUD 2001, 239 and 235, fig. 1; Horsnæs 2002b. Site 060104 sb 251. BMR2374x5. FP 6409. 20 See now Horsnæs 2009. 21 One solidus may be dated to the reign of Justin I. Breitenstein 1944, 78 no. 46; Klindt-Jensen 1957, 234 no. 11 and fig. 198; Horsnæs 2009, 254 no. 37. From Kirseløkkegård. Site 060405 sb 157. FP 220/NM I 19452. 438 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW a b Fig. 6a-b. Smørengegård, FP 6693.14, intentionally cut (a – obverse, b – reverse). a b Fig. 7a-b. Smørengegård, FP 6693.15, intentionally cut (a – obverse, b – reverse). Among the solidi found on Bornholm in general there seems to be a predominance of coins struck at Constantinople, while the mint from Thessaloniki is less well represented, and the remaining eastern mints are not represented at all. Coins from the western mints are less common, but there is a number of coins from Ravenna, Rome and Milan, and a single coin struck at Arles during the reign of Majorianus should be mentioned. There is also a small number – less than 5% of the total number of solidi found on Bornholm – of coins from the various non-Imperial mints and other so-called irregular issues ascribed to various Germanic Kingdoms. Normally the solidi found on Bornholm are very well preserved, though the degree of wear varies. They are rarely pierced or mounted with an eye to be used as jewellery or another forms of decoration. There are, however, at least five solidi that have been intentionally cut (cf Figs 6-8)22, and other 3-4 solidi with traces which may be interpreted as attempts at cutting the coin. Outside Bornholm, this phenomenon is described so far only for two fragments from 22 1. Smørengegård. Site 060305 sb 70. FP 6693.14. BMR 1469x259. Horsnæs 2009, 246 no. 16 and fig. 3:15. 2. Smørengegård. Site 060305 sb 405. FP 6993.15. BMR 1697x158. Horsnæs 2009, 247 no. 18 and fig. 3:19. 3. Dalshøj. Site 060403 sb 135. FP 6364.5 and FP 7727.5. BMR 1639x303. Horsnæs 2009, 253 no. 276 and figs. 3:30-31. 4. Kirseløkkegård. Site 060405 sb 157. FP 220 / NM I 19452 (cf footnote 21). 439 HELLE W. HORSNÆS a b Fig. 8a-b. Dalshøj, FP 6364.5, intentionally cut (a – obverse, b – reverse). Fig. 9. Distribution of denarius finds. e following finds have not been mapped: 10 denarii with provenance “Bornholm”; 2 denarii from “Klemensker parish”; a denarius from Sigård, Klemensker parish, FP 506; 2 recently found denarii from Duegård, Poulsker parish, BMR 3354; a denarius found on the coast south of Åkirkeby, Åker parish, FP 769; 2 recently found denarii from Nygård, Åker parish, BMR 3339 and FP 7271; a recently found denarius from Kastelsgård, Åker parish, BMR 3324 and FP 7088; 3 recently found denarii from Ndr. Mulebygård Øst, Nyker parish, BMR 3227 and FP 6564 and 7106; a denarius from “Rønne”; a denarius from “Kongens Udmark”, Vestermarie parish, FP 224; a recently found denarius from Store Dalbygård, BMR 3145 and FP 7391; 6 recently found denarii from Svanegård, Ibsker parish, BMR 3314 and FP 7429; 2 recently found denarii from Engegård, Ibsker parish, BMR 3408 and FP 7431; a recently found denarius from Højemark, Ibsker Parish, BMR 2755 and FP 7423; a recently found denarius from Fuglesangsagerne, Ibsker parish, BMR 3091 and FP 7425; a denarius from Store Frigård, Østermarie parish, FP 1701; a denarius from Rynsegård, Østermarie parish, FP 676 (Map by Claus Dam, KUAS – Danish Cultural Heritage Agency). 440 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW Fig. 10. Distribution of solidus finds. e following finds have not been mapped: 11 solidi with provenance “Bornholm”; a solidus from “Pedersker”; 6 solidi from “Almindingen”, FP 561; a solidus from “Nylarsker”, FP 1477; a solidus from “Blemmelyng?”, Nylarsker parish, FP 57; a solidus from “Gudhjem”, FP 591; a solidus from “Sylten/Køllegård”, Ibsker parish, FP 676; 2 solidi from “Brændesgård”, Ibsker parish; 8 solidi from “Sylten?”, Ibsker parish; 3 solidi from “near Svaneke”, Ibsker parish; 2 solidi from “Østermarie”; a solidus from Kløvegård, Østermarie parish, FP 260 (Map by Claus Dam, KUAS – Danish Cultural Heritage Agency). the central place at Gudme, Funen (Horsnæs 2009, 257 nos 9-10), and a single solidus from the Czech Republic23. While a considerable number of gold coins from the remaining parts of Denmark have been reworked into jewellery, this applies only to a small number of the Bornholm coins. 12 solidi are pierced and among these there are two examples of a coin where the pierce hole has been sealed again with a miniscule dab of gold24. The recently found Fuglesangs-ager 1 Hoard is quite exceptional25. It comprised 6 solidi and 5 bracteates, all mounted with an eye, as well as gold beads, and must be interpreted as Militký 2005, 1115 and fig. 2.3. e coin is a Barbarian imitation of a eodosius II solidus. 24 From Almindingen Forest, exact find-spot unknown. Breitenstein 1944, 66-68 no. 26.5 fig. 74. FP 561. 25 Axboe 2002; Horsnæs 2001; 2002a; 2009, 239. Site 060304 sb 93. FP 6296. BMR 2649. 23 a grandiose necklace. All elements of the necklace had been dismantled and wrapped into a Late Roman silver plate in which it was found. The coins from the Fuglesangs-ager Hoard are also unusual in themselves, as they were all struck in Western mints (Rome, Ravenna) and there is even a non-Imperial (Gallic) coin.Apart from the Fuglesangs-ager coins only two solidi from Bornholm have been mounted with an eye. The majority of the solidi from Bornholm have been found as hoards, often in connection with other artefacts in gold or (more rarely, however) in silver: bracteates, fibulae, bars. Sites with Roman coin finds Roman coins have been found in ca. 150 sites on Bornholm (Figs 9, 10). Most of the sites are situated as a broad band along the coastline, while the interior of Bornholm has yielded a smaller number of finds – all Pre-Detector 441 HELLE W. HORSNÆS Age finds! This is probably due to the modern vegetation: where the Almindingen woods now cover the surface, detector surveys have not been undertaken. The old finds, however, seem to indicate that the modern distribution map may be misleading. An analysis of the topographical situation of the finds reveals a mixed picture. Many of the coins have been found in close connection to settlement areas, and the solidi have often been found in sites made extraordinary by other finds. Some of the hoards evidently were deposited in connection with contemporary buildings, as is the case of the gold hoards at Fuglesangs-ager (Nielsen 2004, 74-75, fig. 17 no. 1) and Dalshøj (Klindt-Jensen 1957, in part. fig. 24 on p. 29), both sites are part of the Ibsker parish central place, while, for example, the Saltholmgård Hoard consisting of 29 solidi from the 5th and the early 6th c. (Theodosius II to Anastasius) was deposited in a Prehistoric dolmen26. The hoard was found in the western part of the dolmen, under one of the side stones, and the coins were found wrapped up as a small roll. Along with the coins was a bar of silver mixed with gold weighing 299.63 g. The dolmen is situated close to a small brook running west-east to an outlet north of Årsdale.Whether the hoard was deposited in the dolmen, because it represented an ideal hiding place – some distance from the settlement, yet easy to recognize and thereby to retrieve, or because the hoard was meant as a sacrifice (to the forces of nature, the spirits living in the dolmen, or to safeguard the southern border of the Ibsker central place) is difficult if not impossible to ascertain today. The sites are rather varied. In some cases a “site” is the find-spot only of a single coin – and in many cases we have only very general information on circumstances of discovery. In other cases the sites are quite complex, and particularly in the case of many detector finds we are lucky to have very accurate information 26 Breitenstein 1944, 52-58 no. 22. FP 475. Site 060403 sb 68. on the find circumstances. Furthermore, there are great variations in the longevity of the sites yielding Roman coins. Naturally, sites with one or just a few finds cannot be considered in this connection, but among sites with Roman coin finds there are examples both of sites which seem to fade away during the later part of the Iron Ages (e.g. the central place in Ibsker parish) and sites with a marked continuity in the numismatic material into the Medieval Period (e.g. the Agerbygård site). Detector archaeology has focused on the Black Soil sites and the rich detector finds from these sites may have biased our view in favour of settlement archaeology. Some of the older hoard finds, as e.g. the Borresø Hoard27 or the Robbedale Hoard,28 suggest that hoards may have been deposited also some distance from settlement areas. Naturally, this view may change if archaeological investigation in future should reveal the presence of settlement material at these sites. So far Roman coins have been found in only three burials on Bornholm. A denarius struck during the reign of Nerva was discovered in a boat grave at the Slusegård cemetery, grave 33. The burial is dated to the period C1b29. A Trajan denarius was found in Grødbygård, grave 354, dated to C1b. The coin was lying on the chest of the deceased man. It probably was wrapped in a purse or the like30. Finally, a male burial in Rævekulebakke yielded no less that 4 denarii which rested by the pelvis. The coins e hoard was found in the mud of the former drained Borresø (Borre Lake) in Almindingen. It consisted of one or more spiraliform gold finger rings and some denarii, of which 16 were sent to the later National Museum in Copenhagen in 1832. Breitenstein 1944, 33-35 no. 3. Site 060305 sb 154. 28 e 253 denarii were found together with the bottom and some fragments of an “urn” (probably, a ceramic jar) in 1893. Breitenstein 1944, 7-33 no. 2. Site 060304 sb 20. 29 Klindt-Jensen 1978, 35 f.; Andersen et al. 1991, 199 ff. BMR no. 948x345. Site 060203 sb 1. 30 Jørgensen 1987, 81. FP4365. BMR no. 948. Site 060205 sb 205. 27 442 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW were issues of Antoninus Pius (for Diva Faustina Major, RIC 358, after AD 141), Commodus (RIC 71 or 89, AD 184-185 and RIC 159, AD 186-187), and Severus Alexander (RIC 23, AD 223). The date of the burial is again C1b, which means that is it near contemporary with the Severus Alexander coin31. Regarding the burials we can conclude that coins were not among the standard equipment in Iron Age burials, but the three occurrences are of prime importance for the discussion of the date of the arrival of Roman coins to the island. Let me give some examples of the problems involved, based on the material from Bornholm33. The earliest contexts with Roman coins from Bornholm are the three burials with denarii. They have all been dated within the period C1b (210/20-250/60 AD)34. These burials are good evidence that at least some denarii had arrived in Bornholm in phase C1b. But unfortunately these burials are the only well-documented contexts seen from a chronological point of view. The hoards must be dated by the latest element in the hoard. This is easy if the hoard is found intact in situ, but with the situation described above, where we must interpret many of the multiple finds of coins as hoards that have been scattered by the plough, the interpretation of a coin as belonging to a hoard or not may have severe implications for the interpretation of the date of the hoard. The Robbedale Hoard mentioned above seems to be a rare example of a pure denarius hoard found (almost) intact with its container35. Its internal terminus post quem is 194 AD provided by a Septimius Severus coin from 194-195 AD, but the coins from the hoard are quite worn, and it is debatable for how long and under what circumstances the coins should have circulated to Production date – circulation period – deposition date It is a great privilege to work with Roman coins. The important questions on production site and production date are far easier to solve when dealing with coins than when working with most other types of archaeological artefacts. In the 2nd c. AD Roman silver coins were produced exclusively in Rome by the imperial mint32. In many cases the production date can be ascertained to within a year, but even, for example, the undated issues of Trajan and issues struck in honour of the empresses are safely dated within the reign of a given emperor providing us with good post quem dates for the coin finds. The Late Roman solidi are dated more often within the reign of an emperor, but even here dates can sometimes be given with more accuracy. Is it, however, far more difficult to date the deposition of the coins in Bornholm. This problem has severe implications for the discussion on the circulation period of the Roman coins in Barbaricum, as well and on the discussion of the route(s) that the coins travelled north. At present it also seems hard to solve the question about when the Roman coins arrived in Bornholm, but some points may be made. Nielsen 1992, 148 ff. FP 4435. BMR no. 1081. Site 060205 sb 198. I owe the information on the details of the burial to Lene H. Lutz. 31 Other mints were active during the Flavian and Severan dynasties, but their output is relatively small in comparison with the mint in Rome, and in the case of the coin finds from Bornholm these issues are negligible. 33 I will not – in this context – go into details of the scholarly discussions of Roman coin finds in Barbaricum since the days of Sture Bolin’s Fynden av romerske mynt i det fria Germanien (Lund 1926), but only mention the most important contributions of direct relevance for the finds from Bornholm: Joan Fagerlie, Late Roman and Byzantine solidi found in Sweden and Denmark, New York 1967; Lennart Lind, Romerska denarer funna i Sverige, Stockholm 1988; Bursche 1996. e finds from Jutland, Funen and Sealand are discussed in Horsnæs 2010. 34 e absolute dates used for the archaeological phases are taken from Jørgensen et al. 2003. 35 Breitenstein 1944, 33. e coins were found together with the bottom and side of an “urn”. Unfortunately, the “urn”, probably a small ceramic jar, did not survive. 32 443 HELLE W. HORSNÆS achieve this degree of wear. While it is possible that the Robbedale Hoard had been deposited already during period C1b, we cannot exclude a later deposition date. The many denarii from the Smørenge Hoard were found in and around the remains of two small ceramic jars and among this material was also a single Late Roman solidus, struck in Milan during the reign of Anthemius (467472 AD). This led the excavator to conclude that the solidus should be regarded as part of the hoard which, consequently, must have been deposited after 467 AD, well into the Germanic Iron Age, and more than 250 years after the production of the latest denarius in the hoard. The mixture of denarii and solidi is not unknown from other sites; in fact, the Sorte Muld Hoard seems to have yielded a similar case. One of the detector finds from this area was a denarius and a solidus that were found corroded together. Thus having established that while at least some denarii came to Bornholm in period C1b, it is also clear that at least some denarii were not deposited in the ground until the late 5th c. AD. However, some ancient coins were above ground (again?) even during the Viking Age. In 1811 the farmer Jens Jensen came upon hoard of 46 coins, 10 fragments of coins, two silver chains and 28 pieces of hack silver, a total of 437.5 g of silver, on his land at Vellensbygård on southern Bornholm36. The coins were 16 Kufic dirhams37, one English and 29 German coins from the late 10th and early 11th c. Recent detector surveys have identified the original find-spot, and until now an additional 49 coins and some pieces of hack silver have been recovered. The newly found coins are of the same types as the original ones, but there are also a number of somewhat later coins, including the commonly found coins from the Danish Civil War Period (1286-1341) and early Renaissance coins38. Furthermore, there is a single Roman denarius. The denarius is fragmented and it could not be dated precisely. It was cut intentionally twice, and it has furthermore been broken in modern times (Fig. 11)39. The site is a settlement with material dating from ca. 7001100, but no evidence of occupation during the Roman or Early Germanic Iron Age. The distribution map of recent finds shows that the Roman denarius was found amidst the late 10th - early 11th c. coins40. While obviously we cannot exclude entirely that the Roman denarius was an accidental single find, it is quite probable that it was (re-?)deposited as part of the 11th c. hoard, and was regarded as ancient already in the Viking Age. The cutting of the denarius might in itself be an indication of Viking Age re-use. Other examples of Viking Age re-use of Roman coins have been noted, also from Bornholm. Ten denarii have been found during excavations of the Medieval castle Lilleborg41, and two 11th c. hoards, Kongens Udmark (Breitenstein 1944, 38 no. 5; Heijne 2004, 322 find 5.97) and Store Frigård I (Breitenstein 1944, 39 no. 7; Heijne 2004, 331 find 5.135), have both yielded a denarius. No less complex is the Lillegærde Hoard. Here one of Alexander III’s posthumous issues Recent detector surveys have yielded both coins and hack silver: FP 6113 (3 coins), 6346 (5 coins), and 6853 (38 coins). Five coins await registration. BMR 2361. Site 060303 – 207. Coins from the Civil War Period and Early Renaissance coins are quite common finds in all areas (of Bornholm? Denmark?), and they probably have nothing to do with the hoard. 39 e denarius BMR 2361x113; FP 6853.1 40 My warmest thanks go to Finn Ole Nielsen of Bornholms Museum who patiently answered my questions and generously gave me access to the full find-lists and distribution maps of the Vellensbygård site. Finn Ole Nielsen also reminded me of a similar case from Vasagård (unfortunately, the denarius from this deposit is now lost). 41 Mørkholm 1969, suggesting that the coins from the Lilleborg excavations may have been part of the Borresø Hoard; Kromann 1983-84, 110-113 no. 52. 38 Galster 1980, find no. 16; Heijne 2004, find. no. 5.91. FP XI, found 1811. 37 Islamic silver coins, probably from the 10th c. e Kufic coins were not identified and were later returned to the finder. 36 444 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW a b Fig. 11a-b. Fragmented denarius – cut intentionally twice and broken in modern times. Due to extreme wear the obverse and the reverse are not clearly identifiable. was found in another 11th c. hoard of coins and hack silver42. In this case the presence of a Greek coin – in itself a very rare find in a Nordic context – can be explained as an exotic piece that made its way to Bornholm as an antique together with the Kufic dirhams from the Lillegærde Hoard. So far no solidi seem to have been found with Viking Age/Medieval material. Several are from hoards that also contain a number of other objects, mostly in gold. In these cases the non-Roman objects are normally dated to the Early Germanic Iron Age (6th c.). It has been suggested that the majority of the gold deposits were buried in the earlier part of the 6th c., perhaps in 536 AD as religious offerings in response to the “Year Without Summer” described in literary sources (Axboe 2001, 117-123; 2007, 156-158 English summary). Thus the circulation period of the solidi seems to have been much more restricted than the circulation period of the denarii. Aleksander Bursche has argued convincingly that the majority of the Roman denarii 42 AUD 1991, 226, fig. 2.2; Jensen & Kromann 1993; Heijne 2004. find no. 5.126. Site 060405 sb 197. BMR 1513. Coins belonging to this hoard are still being recovered from this site: FP 4989, 5138, 5241, 5460, 5798, 6362, 6857. Nine coins await registration. found in Barbaricum must have left the Empire before or at the latest in the 240s AD, as the denarii are rarely found in hoards with later coins within the Empire (Bursche 1996). Many scholars have maintained that the 2nd c. denarii were used to buy peace for the Romans in periods of trouble, from the Marcomannic Wars onwards. Similarly, the presence of the many solidi has been explained as payments for peace, in this case, made to the Gothic Kingdoms and the Huns. But no matter how we choose to explain the export of coins from the Roman Empire, many important questions can only be addressed from a “Barbarian” point of view and within the framework of a close analysis of the archaeological finds in Barbaricum: what happened after the coins left the Empire? How did they make their way to Bornholm? Which route(s) did they take? How many times did they stop over? Did they all come to Bornholm in the period C1b, or did some arrive as late as the Viking Age? What was the function of the Roman coins in the many different centres of Barbaricum? These questions are still to be answered. The continuous inflow of new detector finds from Bornholm does not make this task easier, but it provides us with a unique archaeological material available for future scrutiny. 445 HELLE W. HORSNÆS References Adamsen, Chr., Lund Hansen, U., Nielsen, F.O. & M. Watt (eds) 2009. Sorte Muld. Wealth, Power and Religion. Rønne. Andersen, S. H., Lind, B. & O. Crumlin-Pedersen 1991. Slusegårdgravpladsen. Gravformer og gravskikke: bådgravene. Jysk Arkæologisk Selskabs Skrifter 14/3. Århus. AUD 1985-2003. Arkæologiske Udgravninger i Danmark 1984-2002. København. Axboe, M. 2001. Amulet pendants and a darkened sun. On the function of the gold bracteates and a possible motivation for the large gold hoards. In: B. Magnus (ed.), Roman gold and the development of the Early Germanic kingdoms, 119-135. KVHAA konferenser 51. Stockholm. – 2002. Sølvkræmmerhuset og balders død – nye brakteatfund fra Bornholm. In: J. Pind et al. (eds.), Drik og du vil leve skønt. Festskrift til Ulla Lund Hansen, 295-303. Publications from the National Museum. Studies for Archaeology and History 7. Copenhagen. – 2007. Brakteatstudier. Nordiske Fortidsminder B/25. København. Breitenstein, N. 1944. De romerske møntfund fra Bornholm. Nordisk Numismatisk Årsskrift 1944, 1-85. Bursche, A. 1996. Later Roman-Barbarian Contacts in Central Europe: Numismatic Evidence. Studien zu Fundmünzen der Antike 11. Berlin. Ethelberg, P. 2000. Skovgårde. Ein Bestattungsplatz mit reichen Frauengräbern des 3. Jhs. n.Chr. auf Seeland. Nordiske Fortidsminder B/19. København. Galster, G. 1980. Vikingetids møntfund fra Bornholm. Nordisk Numismatisk Årsskrift 1977-1978 (1980), 5-246. von Heijne, C. 2004. Särpräglat. Vikingatida och tidligmedeltida myntfynd från Danmark, Skåne, Blekinge och Halland (ca 800-1130). Stockholm Studies in Archaeology 31. Stockholm. Horsnæs H.W. 2001. Et halssmykke fra Bornholm. Nordisk Numismatisk Union Medlemsblad 5, 95-98. – 2002a. New gold hoards from Bornholm with rare types of Valentinian III solidi, Revue Numismatique 158, 131-138. – 2002b. Nye romerfund fra Bornholm. Nordisk Numismatisk Unions Medlemsblad 1-2, 6-8. – 2006a. Roman bronze coins in Barbaricum. Denmark as a case study. In: H.W. Horsnæs & J.Chr. Moesgaard (eds), Single Finds, the Nordic Perspective. Proceedings of the Seminar in Copenhagen 23-24 November 2001, 53-99. Nordisk Numismatisk Årsskrift 2000-2002 (2006). – 2006b. Many coins from one site. In: H.W. Horsnæs & J.Chr. Moesgaard (eds), Single Finds, the Nordic Perspective. Proceedings of the Seminar in Copenhagen 23-24 November 2001, 100-108. Nordisk Numismatisk Årsskrift 2000-2002 (2006). – 2007. Diva Faustina som consul? En hybrid denar fra Bornholm. Nordisk Numismatisk Unions Medlemsblad 2, 47-48. – 2009. Late Roman and Byzantine coins found in Denmark. In: M. Wołoszyn (ed.), Byzantine Coins in Central Europe between the 5th and the 10th century (Congress Polish Academy of Arts and Sciences, Krakow, April 23rd to 26th 2007), 231-270. Moravia Magna seria Polona III. Kraków. – 2010. Crossing boundaries. An analysis of Roman coins in Danish contexts. 1. Jutland, Funen and Sealand. Papers of the National Museum 18. Odense. Jensen, J. S. & A. Kromann 1993. The Hoard of Lillegærde in Bornholm. In: T. Hackens & Gh. Moucharte (eds.), Proceedings of the XIth International Congress Brussels… 1991, 3, 73-80. Louvain-la-Neuve. Jørgensen, L. 1978. En bornholmerpige fra 700-årene – familiegravlæggelser i 1000 år. Nationalmuseets Arbejdsmark 1978, 75-86. Jørgensen, L., Storgaard, B. & L. Gebauer Thomsen (eds.) 2003. The Spoils of Victory. The 446 ROMAN COINS FROM BORNHOLM – A PRELIMINARY OVERVIEW North in the Shadow of the Roman Empire. Copenhagen. Klindt-Jensen, O. 1957. Bornholm i Folkevandringstiden og forudsætningerne i tidlig jernalder. København. – 1978. Slusegårdgravpladsen. Bornholm fra 1. årh. f. til 5. årh. e.v.t.: anlægstyper, planer, plancher. Jysk Arkæologisk Selskabs Skrifter 14/1. Århus. Kromann, A. 1983-1984. Recent Roman coin finds from Denmark. Supplement to Breitenstein and Balling. Nordisk Numismatisk Årsskrift 1983-1984, 59-122. Kromann, A. & M. Watt 1984. Skattefundet fra Smørenge. En nedgravet skat fra Folkevandringstid på Bornholm. Nationalmuseets Arbejdsmark 1984, 29-41. Militký, J. 2005. Finds of Roman and Early Byzantine Gold Coins on the Territory of the Czech Republic. In: C. Alfaro at al. (eds.), XIII Congreso Internacional de Numismática Madrid 2003, Actas, 1113-1124. Madrid. Mørkholm, O. 1969. De romerske mønter fra Lilleborg, Bornholm. Nordisk Numismatisk Unions Medlemsblad 1969, 137-140. Nielsen, S. 1992. Roman denarii in Denmark – an archaeological approach. Nordisk Numismatisk Årsskrift 1987-88 (1992), 147-169. Nielsen, F. O. 2004. Bornholms Museums Antikvariske arbejde 2002-2003. In: Bornholms Museum 2002-2003, 61-84. Bornholms Museumsforening. Rønne. Vennersdorf, M. 2002. Smørenge 2000, Udgravning af en formodet skat fra yngre jernalder. Arkæologisk projekt II: magisterkonferensuddannelsen, Københavns Universitet. København. 447