The Coins in the Grave of King Childeric

Svante Fischer & Lennart Lind 2015. The Coins in the Grave of King Childeric. This article contextualizes some one hundred midto late 5 th century solidi and two hundred silver coins found in the grave of King Childeric in Tournai, Belgium. We argue that the coins in the grave must have been assembled for the specific purpose of the burial rite and that some of the participants in the burial rite were allowed to look at the coins before the grave was sealed. We argue that they were capable of identifying the various coins because they were literate and familiar with Roman iconography. It follows that the solidus hoard together with the other coins is a meaningful composition that has been manipulated for ideological purposes by Clovis himself. The coins must hence be explained in a manner that considers Clovis’ ideological motives, as the grave and its contents run contrary to all usual explanations.


Introduction
Childeric's grave in Tournai was probably built in AD 481-482 at the request of his son Clovis.The grave was discovered on May 27, 1653, and then followed an unscientific excavation of the burial monument.The latter enclosed a rich assembly of grave goods, including coins.The Childeric grave is the only known inhumation burial with a mixed gold/silver coin-hoard of three hundred coins covering five centuries and thus constitutes an anomaly beyond all comparative estimates.We argue that it was Clovis who decided what was to be put into his father's grave.
The unusual find combination of coins from a funerary context was first described and published by Chiflet with the help of his son (Chifletius 1655). 3 In 1655, Jean Chiflet was a Jesuit scholar in his fifties.Although absent from the excavation of the grave, Chiflet was able to see and illustrate the finds after they had been removed from the burial context, and subsequently cleaned.The quality of Chiflet's illustrations shows a considerable knowledge of Roman numismatic iconography.He was undoubtedly familiar with the names and the chronological order of the Roman emperors just as he was familiar with a number of Roman coin hoards discovered in his own time.Chiflet had no trouble understanding the occasionally quite stylized representation of the shield on the reverse of the solidi.
In this paper, the Chiflet publication and its description of the coins are compared to a variety of archaeological contexts, especially similar find combinations including late 5 th century solidi and denarii, notably the Vedrin hoard in Namur, Belgium, but also hoards found in settlements on the Southeast Scandinavian islands from Bornholm in Denmark, and Gotland, Helgö and Öland in Sweden. 4We argue that the burial of Childeric was a key event that merged many different traditions, but also highlighted certain 4 The research for this paper has been generously financed by the Royal Swedish Academy of Antiquities, History and Letters, KVHAA, through its Western Europe scholarship and the Enbom Foundation.We wish to thank Birgit Arrhenius, Ilona Bede, Frands Herschend, Michel Kazanski, Ulf Näsman, Patrick Périn, Dieter Quast, and Ulla Westermark.We are also indebted to the helpful staff of the KMK and the KVHAA library.
Fig. 1 Reconstruction of the burial site in St. Brice, Tournai. (After Brulet 1990). 1 = The main tumulus. 2 = Horse burials.http://clamator.its.uu.se/uploader/92/JAAH14_Fig_1.pdfnotions of power, especially the ability to understand and interpret specific attributes of legitimate power, such as texts.McCormick (1989, 156) has discussed Gregory's of Tours account of Clovis' triumphant parade into the episcopal see of Tours in AD 508 in terms of a deliberate imitation of Byzantine public ritual. 5e argue that one must differentiate between those invited to inspect the burial closely, and those who were not.We argue that a small part of the literate élite present at the burial was allowed to see the grave goods up close, and to read specific texts on certain key objects.This argument is based on the presence of evidence in favor of a literate audience.There are more than 300 preserved letters between people in Gaul written in the period AD 420-500 (Mathisen 1993, 23-24;2001).Ruricius of Limoges alone is known to have written at least 82 different letters.Typically, people in Gaul wrote to others in Gaul and not to people in Italy, for instance.Both Childeric and Clovis are likely to have been part of this corresponding network.One must begin the account of the coins of the Childeric grave by reiterating two dry facts that have been quite correctly stressed by the historian Halsall (2010, 187). 6First, it is not the buried individual who decides what grave goods are to accompany him or her into the afterlife.Rather, this is something ultimately decided by the descendants.In the case of the grave of Childeric, it is rather easy to pinpoint the responsible descendant.Second, there is no point in depositing precious grave goods unless they are arranged in a meaningful order that may be appreciated by those partaking in the burial rite.Therefore, we argue that Clovis' burial of his father was an important transitory event.
It materialized once and for all the aspirations of the Germanic successor kingdoms in northwestern Europe -the acquisition of dynastic legitimacy in the eyes of both the Roman Empire and nomadic and Germanic warlords.
Clovis apparently succeeded his father to the throne without any major interference from neither relatives nor unrelated rivals.The opulent burial of his predecessor was the first real manifestation of power in the reign of the young Clovis.Later he would go on to conquer Gaul, convert to Catholicism after a victory over the Alemanni, defeat the Visigoths at Vouillé in AD 507 and become an honorary Roman consul during the reign of the eastern emperor Anastasius.We can thus establish that Clovis is the single individual who must have supervised the deposition of the coins in his father's chamber grave in Tournai.
But the coins given by Clovis to Childeric to guard in his afterlife puzzle us.Clovis left a riddle behind.He deposited coins stretching over five centuries, from the Roman Republic in the 1 st century BC down to the eastern emperor Zeno (476)(477)(478)(479)(480)(481)(482)(483)(484)(485)(486)(487)(488)(489)(490)(491).But the coins have obviously been arranged in some form of meaningful order.Why?What were Clovis' motives?This can only be discerned if we assess all the major grave goods and burial structures in the grave complex through the framework of the burial rite.

The Burial Rite
How can we begin to understand the nature of the burial rite?It would appear reasonable to assume that a burial rite has a beginning and an end.For an account of the interpretation of the Merovingian burial rite within archaeology, see Effros (2002;2003).There may be words spoken aloud along with a procession of mourners.One could further surmise that the rite is intended to relate to important events or acts in the real life of the deceased throughout its necessary duration.It would probably take a little more than one full hour's time to recapitulate Childeric's life in front of an audience.
We know very little of Childeric's life.The eulogy delivered at his burial would have been useful information indeed.Similarly, a preserved letter describing the event would be very helpful.But one may find comfort in that we do know of the acts of one of his contemporaries, Theoderic, king of the Visigoths in Aquitaine.In a letter to his brother-in-law Agricola, Sidonius Apollinaris details the daily routines of Theoderic in AD 454 presumably at a residence in or around Toulouse. 7We can thus see that the king first sits on his throne for one hour or two.He then inspects his treasure chamber.After this, he usually goes to inspect his stables.This structured routine bears a very strong resemblance to the composition of Childeric's grave in terms of delineating action in time and space, (see fig 1).At the center is Childeric in his grave.His long hair was probably neatly combed, and he was in all likelihood wearing a chlamys adorned by gold and garnet bees (see fig. 8), clasped together by a Roman official's fibula (see fig. 8b), his personal signet ring on his finger (see figs 4-4b), kolbenarmring around his wrist (see fig. 6), and a crystal-ball scepter in his hand (see fig. 7).Next to him were his coins (see figs 2-3), tucked away in purses (see fig. 5), and his other regalia and weapons (Böhner 1981, 453;Kazanski and Périn 1988). 8Outside his grave were a vast number of horses, a stable for the after-life (Brulet 1990).Clovis probably made sure that the right people were invited to the staged event; Gallo-Roman church leaders and imperial functionaries along with highranking Frankish nobles are likely to have participated in the burial rite.The names of those attending besides Clovis will probably remain unknown to us.Still, it is tempting to suggest a few people who ought to have been invited, notably bishop Principius of Soissons, his brother Remigius, bishop of Reims, the Roman warlord Syagrius, the owners of the Vedrin hoard, but also men such as the warrior chieftain possibly named HEVA who was buried in Pouan (Aube), cf.Kazanski (2003).For the event to have the intended effect, it is possible that the participants were required, requested or invited, all depending on rank, by Clovis to walk past Childeric on his lit de parade.A select few would then proceed to peruse the main objects of his treasure chamber, after which they were all free to venture further away to look at the stables.When everything was found to be in good order by the select few, all participants could congratulate Clovis to a job well done.Presumably, he, in turn, could now ask the most prominent among the guests to be seated at his royal table to eat and drink.

Reading the Solidus Hoard
Pictura est laicorum litteratura.Hence, scholars unfamiliar with the field of epigraphy, both Latin and runic, often appear all too keen to embrace the socalled "topos of the illiterate warrior".As shown by Grundmann (1958), the topos belonged to a set of ideas that was typically supported by the late 19 th century German national romanticists.Adherents of the topos generally argue, without evidence, that barbarian kings were always illiterate because they were virtuous noble savages.(For a discussion of this problematic in relation to Theoderic the Great, see Fischer 2013).While this old-fashioned topos still retains a non-negligible influence, it is inappropriate as a scientific method, because it has to argue against evidence without being able to present any of its own.Believing that barbarians were illiterate is one thing, proving it in the face of the corpus of preserved texts is simply not a realistic option.Therefore, accepting that late 5 th century texts were there to be read, that there was a literate audience, albeit restricted, is to proceed in accordance with Ockham's razor.
It is beyond all reasonable doubt that the leading members of the late 5 th century élite in Gaul were able to read Latin capitals (Mathisen 1993(Mathisen , 2001)).They were also familiar with the iconography of Roman coinage.It would not have been very difficult for anybody used to handling solidi to tell a western type obverse with a diademed profile apart from an eastern obverse with a cuirassed and helmeted three-quarter profile.Similarly, the vota issues, with their very particular iconography of an emperor robed in checker-patterned chlamys with a diademed profile, were meant to be understood as rarities from the very onset, see below.All native Latin speakers who were literate, such as Remigius of Reims or his brother Principius of Soissons, would have had no trouble reading a coin legend DN LEO IMP PF AVGGG, nor would a Roman military commander such as Syagrius have any trouble understanding what VOT X meant.Principius of Soissons was one of the most prolific correspondents of his day, and his letters to Sidonius Apollinaris are preserved and make for a very interesting read.
The undeniable fact that Childeric himself carried a signet ring with a cuirassed representation of himself (clearly an imitation of the image of the emperor on the solidus coinage), and a legend that sported both his own name and his title in a Latin capital legend CHILDIRICI REGIS suggests that members of the élite were meant to understand what both the image and the letters meant (cf.Salaün et al. 2008), see figs.4-4b.
If everyone was supposedly illiterate, why were there legends on objects such as finger rings, coins and tombstones, then?It must be emphasized that there is simply no evidence that Late Roman Gaul in the last quarter of the 5 th century AD was run by an illiterate élite -on the contrary.The future emperor Avitus (AD 455-456) even served as a teacher to the Visigothic king Theoderic in the early 5 th century (Mathisen 2001).The early 6 th century grave of Krefeld-Gellep 1782 is an interesting comparison, being a context well known to most archaeologists and numismatists in the field (Pirling 1964).Why have a Germanic name Arpvar in Latin text with poetic meter ARPVAR ERAT (F) ELEX UNDIQUE PRE (Arpvar erat felix undique praecelsus), that is, "Arpvar was happy, and respected everywhere" on a pitcher if nobody could read the poem praising Arpvar out aloud?
The fact is that barbarians and Romans alike wore signet rings with titles, names and monograms, and wrote letters and poems to each other.They exchanged money, coins with standardized and meaningful texts.It is of great significance that there was a long tradition in Belgica Secunda of barbarians being brought into Roman society (rather than the Romans giving up Roman territory altogether as in Dacia or the Agri Decumates), something that also explains the success of the Merovingian dynasty in this region.Barbarians were settled in northern Gaul already in the late 3 rd century AD during the Tetrarchy, and there is further evidence in the Notitia Dignitatum.By the 4 th century, Magnentius (AD 350-353) a descendant of barbarians residing in Gaul had already managed to become emperor.He was born in Amiens in AD 303 to a Frankish mother and a Breton father.While Magnentius was frowned upon for being a usurper, no Roman historian ever accused him of being illiterate.Less than a half-century later, the Frank Arbogast became the supreme commander of the western army in AD 392, an onerous duty that he nevertheless managed to fulfil for many years.This was probably because he knew how to read.
Accepting that literacy played a fundamental role in public ritual begs further questions.This is where we must begin to ask what was made explicit in the composition of the coin assemblies in the grave.The Childeric solidus hoard itself is unusual in many ways.First, it is the single largest solidus hoard ever found in a grave.Second, it has a distinct composition that puts it into a clearly defined group of hoards that are found in three locations: I) northern Gaul, II) southeastern Scandinavia, III) northern Italy (Fischer 2011;2014b;Fischer et al. 2011).Third, the solidus hoard is accompanied by a denarius hoard and a siliqua, at a time when denarii are relatively difficult to come by inside the Roman Empire and the siliquae have begun to run scarce, although the latter do occur as grave goods in the early 5 th century Germanic chamber graves 205 and 211 in Vron (Somme), see Fischer et al. (2009).Fourth, the other burial goods included chlamys, scepter and a signet ring, indicating the desire to manifest the high status of the deceased in relation to the Roman Empire (Quast 2010;2011, Toth 2012)).Fifth, there is a more barbarian assembly of a kolbenarmring, weapons and horses that is reminiscent of princely burials in Barbaricum, showing an eclectic mixture of Late Roman, Germanic and nomadic attributes of power (Werner 1980;Kazanski and Périn 1988;Brulet 1990).
The purpose of the coins, then, was to display political and ideological legitimacy.Access to solidi would not have been very difficult to a successful warlord like Childeric.There would have been plenty to choose from in a royal treasure chamber at the time (cf.Gasparri 2004;Hardt 2004).Thus, the choice of what solidi should be included would have presented no difficulty to Clovis.He and his closest advisors were probably well aware of who the different 5 th century emperors were and if they also were considered legitimate by the eastern emperor.He also knew that current eastern coinage carried a higher standard weight than western solidi during the third quarter of the 5 th century AD.
Therefore, all the "bad" solidi Childeric had received as a warlord in Gaul and northern Italy during the collapse of the western Empire have, with a few notable exceptions, been removed from the assembly.By contrast, the rare specimen of the western consular solidi of the legitimate and eventually senior emperor Valentinian III in AD 435 (see fig. 2, coin 3), the eastern issue struck in Valentinian's name by the new junior emperor Marcian in AD 452 (see fig. 2, coin 4) and that of the new régime of the senior emperor Leo I in AD 457 (see fig. 2, coin 6) were probably some of the major "good" showcases of the purse in Childeric's lap.
Clovis could count on that the prominent attendants would appreciate this window-dressing of the treasure inventory.After all, they had just like Childeric, once served under a variety of semi-legitimate forms of western government, with substantial issues of "bad", illegitimate and/or underweight solidus coinage greasing unwashed palms.They all knew less about the turbulent political affairs in Constantinople in AD 474-477.Clovis included all these "good" types for good measure: Basiliscus, Basiliscus and Marcus, Zeno, Zeno and Leo Caesar (see fig. 2, coins 8, 10, 11, 12).He also included a western solidus struck in AD 474 for Julius Nepos, Zeno's appointee in the west (see fig. 2, coin 9), but there is no coinage for Zeno struck by Odoacer in Ravenna and Milan after AD 476.By contrast, these later "bad" post-AD 476 issues are a striking feature of the Vedrin and Helgö hoards, which enabled Lallemand (1965) to track most of the die-types (Fischer 2014b).

The Comparative Context of Late 5th Century Solidus Hoards
A survey of previous research shows that the scholarly interpretation of the solidus hoard has largely focused on the origin of the hoard in order to explain its ideological function, despite the fact that most scholars have qualified the assembly of burial goods as an independent ideological demonstration.The numismatist Lallemand (1965, 117) and the archaeologist Werner (1980) incorrectly perceived the Childeric solidus hoard as a direct payment from the eastern emperor in Constantinople to Childeric. 9The scholars concluded that the contents reflected solidus circulation with the Late Roman Empire, based on their current knowledge at that time, and hence the interpretation was rather simplistic by our standards.Later, Böhner (1981, 454) more or less uncritically accepted this stance.The solidus hoard was thus perceived as further evidence of an eastern origin for all the burial goods.But much of this reasoning appears to be inaccurate given that subsequent research has been forced into an impasse of sorts.
In stark contrast to the eastern theories, Kazanski and Périn (1996, 203-209) have argued that the cloisonné work may well be of Italian origin and should be explained in the light of Childeric's connection to the leading warlords of the western Empire, notably Ricimer and Odoacer. 10While incorrectly stating that all solidi but one in the Childeric hoard are of eastern origin, Toth (2012, 280 n.10, 283) is probably right to argue that the fibula in the grave is of eastern origin, and does not belong to the same group as the one in the Reggio Emilia hoard that hails from a western workshop.We must therefore abandon a general eastern hypothesis and seek a new synthetic explanation in which the composition of solidus coin hoards play a constituent role.A new synthesis will require the use of further comparative numismatic data.Unfortunately, Lallemand's view was followed by later numismatic scholars, especially Radnoti- Alföldi and Stribrny (1998) without any additional comparison of the composition of the hoard to a larger corpus.Nor was the RIC X publication of Kent (1994) with its improved typology employed to deconstruct these assumptions, which lead to further erroneous conclusions.A case in point: Radnoti-Alföldi and Stribrny (1998, 44, note 9) appear completely unaware that certain solidus issues celebrating the tenth vota, VOT X, of Valentinian III were struck in both Rome and Ravenna in AD 435 (RIC X 2032-34 in Rome andRIC X 2035-36 in Ravenna).They thus mistakenly claim that Chiflet was in error when he illustrated the correct legend RV for a RIC X 2035-36, (see fig. 2, coin 3).But Chiflet's illustration is quite in accordance with the fact that precisely such issues were struck in both mints, with some transferred Rome dies of RIC X 2032 even having their legend RM recut to RV, as was duly demonstrated already by Kent (1994, 369, Plate 50), see Fischer et al. (2011, 198).An important exception to this general pattern of negligence in regard to the solidi of Childeric is that of Grierson and Mays (1992) who quite correctly pointed out singular peculiarities in the hoard.
In his article on burial contexts with denarii in Gaul, Martin (2004) clearly adhered to the proposition that most of Childeric's grave goods derive from barbarian contexts connected to the eastern Empire.As noted by Martin (2009, 13), there is a substantial lacuna between the last solidus hoards of the early 5 th century in northern Gaul and that of Childeric.Therefore, Martin had already advanced the hypothesis that the solidi reached Childeric via Thuringia (Martin 2004, 260).This is difficult to accept for a number of reasons.Above all, there are no late 5 th century solidus hoards in Thuringia that can be employed to substantiate this claim.The closest parallel is from Biesenbrow, Brandenburg, and belongs to the mid-6 th century.11There are a number of further arguments against these eastern and Thuringian theories.
First, coin hoards are assembled and deposited during all periods due to very specific circumstances regardless of warfare (Sarvas 1968;1970).12Second, coin hoards are related to each other as they reflect the total output and circulation (Thordeman's Law, cf. Thordeman 1949).Third, the weight/frequency ratio of coin hoards falls accordingly in time and space (Gresham's Law).This means that a very vast corpus of comparative material is needed to account for any specific hoard (cf.Fischer 2014b, a recent analysis of 49 different contexts to interpret the solidus hoard of Casa delle Vestali).
In addition, a number of numismatists have correctly pointed to two further facts.Solidus hoards found outside the Roman Empire, in Poland and Scandinavia in particular, cannot be the result of commerce, but of tributes or payments to military units (cf.Herschend 1980;Kyhlberg 1986;Fischer 2005;2008;2014a;2014b;Guest 2008;Ciołek 2009). 13The composition of the Childeric solidus hoard has more in common with the Italian hoards of Reggio Emilia and Zeccone, along with Radostowo in Poland, but also with Belgium and Scandinavia, than with the mint of Constantinople itself (Grierson and Mays 1992, 288-291).
The Vedrin hoard of 69 solidi was discovered already in 1920.This hoard is unquestionably the most important key to the comprehension of the Childeric hoard.It merits considerable attention as a comparison, see table I. Lallemand (1965) published the hoard many years later.The publication included a groundbreaking study of die-identities for solidi struck in the name of Julius Nepos and western issues for Zeno.Two years later, Fagerlie (1967) published the then current Danish and Swedish solidus catalogue, highlighting the unusually high frequency of die-identical coins in the Scandinavian material.With the aid of Fagerlie's publication, Westermark, director of the KMK in Stockholm, studied Lallemand's 1965 publication in great detail and marked all die-identities between Vedrin and Scandinavian hoards with a led pencil in a copy preserved in the library of the KVHAA.Westermark (1980;1983) also published all new Swedish finds and re-discovered solidi in the KMK, linking them to other hoards .Some ten years later, Kent published the tenth volume of the RIC, a typology that was to a considerable extent built on his studies of Swedish solidus finds kept at the KMK (Kent 1994).In 2009, Fischer read Lallemand's article at the KVHAA library and rediscovered Westermark's KMK pencil notes.Fischer was also able to acquire archaeologist Malmer's annotated personal copy of Fagerlie's monograph, from which Malmer (1977) went on to write his study on the comparative chronology of solidi and bracteates.After building the database LEO with over 7,600 solidi, Fischer (2011;2014b) could then proceed to present a case for the interconnectedness of all major solidus hoards in Belgium, Scandinavia and Italy using the RIC X typology and Fagerlie's catalogue supplemented by Westermark (1980;1983) and Malmer (1977), see fi g. 9. LEO consists of 1,683 identifi ed issues of RIC X, and an additional thousand that can be roughly identifi ed according the RIC X.By the same token, it is relatively easy to add to this sum the 1,443 solidi of the Szíkancs hoard (Biro-Sey 1976; Guest 2008;Kolníková and Pieta 2009).We can thus see the relative frequency of coin types in fi nd categories ranging from random fi nds to hoards with hundreds of solidi.Th e Childeric hoard has coin types that match at least 2,491 solidi in LEO.Th is allows for a rather certain estimate of just how typical the composition of the Childeric hoard was at the time of its deposition, see table II.It is obvious that Vedrin is the closest hoard in time and space, and also contains a single denarius struck for Antoninus Pius after AD 141. 14 In terms of its solidus composition, Vedrin spans over a long time from Magnus Maximus in AD 383 to the reign of Anastasius (AD 491-518), extending into the early 6 th century.Two of the earlier coins have a local connection of usurpers gaining access to the Trier mint, having been struck there in AD 383 and 407 respectively.Vedrin is also well connected to the contemporary Scandinavian hoards such as Åby on Öland, Helgö in mainland Sweden, and Botes on Gotland by means of die-identities of both eastern and western issues.The Vedrin hoard contains almost all the pertinent solidus issues that are so obviously missing in the Childeric hoard, even rare specimens of Petronius Maximus (AD 455), Avitus (AD 455-456) and Glycerius.But the Vedrin hoard is also a key to the understanding of the Italian mints under Odoacer after AD 476.Vedrin serves as a nodal point for a number of different dieidentities found on Öland that then branches out through the hoards of Åby and Björnhovda, but it also extends to Gotland and the Helgö hoard in the Mälar Valley on the Swedish mainland.However, the Vedrin hoard contains two RIC X 2831 that share a reverse die-link and a RIC X 2809 (Lallemand 1965, pl. IV, nos. 39-40, 41;Ungaro 1985, 83, no. 15).This means that three out of 69 coins in Vedrin are related to 323 out of 397 coins in the Casa delle Vestali hoard via the imperial mint in Rome.Another possible link between the Vedrin and Casa delle Vestali hoards is that the Vedrin hoard contained coinage from the solidus stock of the military state apparatus in Italy (Lallemand 1965, pl. IV, nos. 42, 43).The people responsible for assembling the Vedrin hoard are thus likely to have been in close contact with the highest echelons of the western Empire around AD 467-476 -precisely the background that would apply to Childeric.The statistical likelihood that the solidus hoard in the grave of Childeric could be a random composition is minimal in the light of the comparative evidence from other contemporary solidus hoards (cf.Fischer 2014b, 116, Table III).Given the geographical location of the hoard and the access to solidus coinage in the region after AD 476, it would be extremely difficult to explain how Clovis accidentally could have grabbed precisely these coins by chance out of some treasure chest.But a comparison with other hoards containing similar coins permits other conclusions.Tables I-II show that the Childeric hoard has very little in common with eastern hoards such as Bína in Slovakia, Szikancs in Hungary or Abrittus in Bulgaria, but looks rather similar to Italian, Polish and Scandinavian hoards.
This brings us back to the actual nature of the majority of the solidi in the Childeric hoard (see fig. 2, coin 7).They form the bulk of good solidi, mostly struck in AD 471-473, and taken out of the eastern treasury to be used in Italy during the late reign of Anthemius and its chaotic aftermath.The largest part of the Childeric hoard must have been assembled in northern Italy in the mid-470's AD.Childeric probably received the money there before or shortly after Odoacer gained control over Ravenna.A key to the understanding of political events in Italy and the financial collapse of the western Empire is the composition of solidus hoards and their respective frequency of western dieidentities.This is precisely the period when Childeric disappears in historical records.Coincidentally, the origin of the coinage struck in Constantinople can easily be explained in the light of the comparative evidence within exactly the same time frame.After AD 476, the western government had small assets of financial capital that must have dried up at a fairly rapid pace; the meager funds were used by Odoacer to pay off mercenaries returning back to northern Gaul and southern Scandinavia.Odoacer would soon try to remedy this downsized situation by issuing an interconnected series of solidi in the name of Zeno after things had settled down in the east, realizing that the east was no longer interested in financing any major actions within western government.
There is thus a difference in hoarding patterns between various regional hoard groups around AD 476.In particular, the Reggio Emilia hoard and the Esquiline hoard in Italy are testimony to the Roman gold hemorrhage after AD 476, while the closely-knit hoards of Vedrin and Öland show the flux of the financial capital prior to AD 476.In the interconnected hoards of the earlier group we also find that for some western emperors to be present there must be certain preceding series of western coins.In particular certain issues of Libius Severus' and Anthemius' coinage condition the presence of solidi for Glycerius.This is the case in Vedrin in Belgium, but also in San Mamiliano, Bostorp, Sandby nr 10 in Högby, Algutsrum on Öland, and Saltholm on Bornholm.This linear relationship between various late 5 th century western solidi is very important.Already Kyhlberg (1986) discussed the phenomenon at length.
Given this result, the composition of the Childeric hoard becomes a very clear-cut anomaly.If the hoard was deposited in AD 481/482, we would have expected the hoard to position itself in between the first group that is interconnected by die-identities prior to AD 476 and the second group with new coinage struck after AD 476, such as Reggio Emilia and the Esquiline.This is not the case however.Instead, the hoard contains elements of the first group's earliest coinage but not the normal western bulk from the 450's to the early 470's AD that usually accompanies the kind of eastern coinage of the same period (see the catalogues of Panvini Rosati (1985); Kyhlberg (1986); Grierson and Mays (1992); Kent (1994); Fischer (2014b)).
The solidi in the Childeric hoard appear to have been gathered from at least two, perhaps three separate sources.One early western source contained specific western payments of the tenth vota of Valentinian III of 435.These solidi are very easy to recognize due to their peculiar iconography and coin legends and do not occur as single finds in northern Gaul at all, see the catalogue of Martin 2009).For an explanation of the vota in the 5 th century western empire, see Burgess (1988), Gillett (2001).The other source, including the two solidi that remain today, consists of a bulk of more recent of the very frequent issues RIC X 605 and 630 struck by Leo I in AD 466-468 and 471-473 and exported en masse to Italy to support the reign of Anthemius, slightly augmented by earlier coinage for Marcian of types RIC X 506-51.There were at least three consular solidi in the hoard, (see tables I-II, fig.2, coins 3, 1, 6).The first is from AD 435 for Valentinian III in Ravenna (for a discussion of this type, see Fischer et al. 2011), the second is a very frequent issue from AD 441 for Theodosius II in Constantinople and the third a relatively infrequent issue for Leo I struck in AD 457 in Thessalonica.The latter issue is also the last consular issue to appear in the east for the remainder of the 5 th century AD.People were clearly aware that this was a rarity: there is also the single looped consular solidus RIC X 623 on Gotland, a region where looped bracteates and regular solidi are aplenty, but where this represents a unicum (the specimen is, in fact, the plate coin of the RIC X catalogue, cf.Kent 1994;Fischer 2014a, 156).These very odd appearances made precisely these solidi stand out, and there still is no obvious satisfactory answer as to how they had ended up in such peculiar circumstances.
Given the presence of the relatively wide range of western solidi in the hoard, it is quite likely that Childeric may once have had access to coinage struck in the name of Majorian (AD 456-461), Libius Severus (AD 461-465), Anthemius (AD 467-472) and maybe even Glycerius (AD 473-474) -either genuine issues or as Visigothic imitations.Many events seem to have been conditioned by payments in return for military services.In AD 458 Aegidius and Childeric aided the illegitimate western emperor Majorian by pushing the Burgundians out of Lyons and then opening up the communications from northern Gaul down to Arles.In AD 463, during the reign of the illegitimate 20 western emperor Libius Severus, Childeric assisted Comes Paulus of Angers in fighting the Visigoths at Orléans.In AD 465 he laid siege to Paris.He then fought Saxons and Bretons in AD 469 during the tumultuous reign of Anthemius, but also attacked the Alamanni.He must have been paid at some point during this period.But with what coinage and struck by whom?
The oldest solidus in the Childeric hoard dates to c. AD 431, and the youngest possibly to AD 477, giving a hoarding range of some 46 years (see tables I-II, fig.2, coins 2, 12 ).But the hoard is void of most western and barbaric issues between AD 435 and 474, some 40 years.Why is this?The answer is twofold.First of all, the hoard appears to have been cleaned.It does not contain any issue emanating from any of the illegitimate western emperors, and the abundant western coinage of Anthemius is absent (cf.Fischer 2014b).Many western emperors are missing although they should have been there if the hoard reflected normal circulation patterns, that is specific payments that were hoarded together over time.It could potentially also have included Visigothic imitations, as found in hoards in England, Italy and Scandinavia.
Why is the Childeric hoard composed as it is in AD 481/482?A possible answer to this question is that after AD 476, things had changed.There was no longer a western emperor.A few years earlier, in the late 460's AD, Anthemius struck quite a substantial amount of solidi (cf.Ungaro 1985;Fischer 2014b).His fall in AD 472 may have caused him to be considered illegitimate.Another possible explanation is that Anthemius' reign in the west AD 467-472 was no longer seen as legitimate by adherents of Zeno after the failed rebellions of Anthemius' two sons in Constantinople in AD 479.There are no nonlegitimate western emperors or usurpers in the Childeric hoard, although their solidus coinage is present in the nearby Vedrin hoard and in the Scandinavian hoards in general, although not in all hoards in Pomerania.The common distribution pattern shared by Scandinavia and Vedrin means that there must have been a very conscious selection of solidi in the Childeric hoard prior to the deposit.This is also precisely why the attempts of Halsall (2010) to alter the chronology of the burial appear ill founded.The two preserved coins in the Childeric hoard struck for Leo I are RIC X 630, these were minted in AD 471-473 and do not appear in the Case delle Vestali hoard, deposited in AD 472, (see fig. 2, coin 7).Had Childeric died before AD 474-476, there would not have been any eastern solidi for Zeno in the hoard, had he lived past the 480's AD there would have been western solidi struck in Milan in the hoard, as in Vedrin and Helgö.

The Silver Coins
The denarii and the siliqua are a different matter altogether.Their ideological role was in all likelihood of a different kind than that of the solidi.Clovis and the other dignitaries inspecting the lit de parade are all unlikely to have been able to identify the imperial personae depicted on the silver coinage.The idea of mixing different coinage in a grave is very uncommon.The denarii themselves are not unusual, though, although they were difficult to come by in Gaul at the time (cf.Martin 2009).But in this context the silver coins have a collective configurative role in the burial rite.They are there in bulk as an ideological entity.They must be interpreted as one single symbol of opulence in the form of Imperial silver in Barbarian custody, as had been the tradition in leading affinities outside the Empire for many generations already (see fig. 10).The combination of the two coin types was ample proof that Clovis, like his father before him, was loyal to the eastern emperor while simultaneously well connected to the top political players outside the Empire.
According to Chiflet there were more than 200 silver coins in the grave: "Nummi argentei duceni, eoque amplius" (Chifletius 1655, 270), but he himself saw only 42 specimens.41 of these seem to have been Roman denarii.The remaining coin was a siliqua of the mid-4 th century, with the portrait of Constantius II (AD 337-361).Although the main part of these more than 200 silver coins, which were worn (cf.Chifletius 1655, 38), probably consisted of just denarii, there may have been more than one 4 th century coin.Of this, however, we will never know for certain.Among the 41 supposed denarii, there was one coin, labeled consularis by Chiflet, with no further information added.This was most certainly a denarius of the Roman Republic.The other 40 were of the Imperial period but mostly determined only according to portrait of emperors and empresses (Chifletius 1655, 270).Whether one chooses the number 200 or 41, it is the largest number of denarii found in a grave.The 42 silver coins can be listed as follows in table III.
Drawings were made of four coins, all pierced (Chifletius 1655, 271), and these are the only ones among the 42 possible to determine more closely than to portrait (see fig. 3).One of these four coins is the single 4 th century siliqua, the other three are denarii, of the 2 nd century AD.They are of common types, Hadrian RIC 244(d), AD 134-138, Antoninus Pius RIC 181, AD 148-149 -a pierced denarius of the same type has been found on Gotland (Lind 1981, 53, No. 61a:42) -and Lucius Verus RIC 463, AD 161.The coin of Constantius II can be dated to AD 351-355.As to the denarii, the Republican coin must have been struck before 30 BC, the Imperial ones from AD 50 or rather 54 (or perhaps 64) to 217.The coin with the portrait of Nero, the earliest Imperial one in the grave, may belong to the latest part of the reign of Claudius (AD 41-54), when coins with Nero's likeness were struck.These, however, are very rare, compared to those issued during Nero's own reign (AD 54-68), hence 54 is more likely than 50.As a matter of fact, the overwhelming majority of denarii with Nero's portrait in finds belong to the last four years of his reign, and it is more likely than not that the coin in the grave belonged to the period 64-68 rather than that of 54-64.
The two youngest coins in the list, those of Julia Domna and Caracalla, may both have been issued before AD 200, but may just as well have left the mint as late as in AD 217.Coins with the portrait of Julia Domna seem to have been struck from AD 193 to 217 and coins with the portrait of Caracalla from AD 196 onwards.Nevertheless, most of the silver coins were 300 years old or more when put into the grave.Where did they come from?Were denarii of the first two or three centuries of our era still at large in Gaul in the late 5 th century?This set of questions appears more complicated than is usually assumed.
It must be stressed that our knowledge of monetary conditions and coin circulation in the Roman Empire is faulty.The absence of clear information on the subject in contemporary written sources, makes all conclusions based on coin finds and general economic theories, however ingenious and subtle, such as Hopkins (1980;2002), Schubert (1992) and Duncan-Jones (1994), entirely hypothetical (cf.Lind 1993Lind , 2006Lind and 2009)).But we know that coins at times did circulate for a very long time in the Roman Empire.A good example made by archaeologists in 1989, is the find of Vienna Rennweg 44, deposited in AD 138 or later.It consisted of 853 Republican denarii, 102 of them belonging to the 2 nd century BC, and 408 Imperial silver coins, mostly denarii and mostly of the period AD 64-138 (cf.above about the reform under Nero).The latest coin in this hoard was almost 300 years younger than the earliest (Dembski and Zavadil 2004).There are other examples.
Most of the denarii in Childeric's grave were issued before AD 193, and were as such presumably "good"silver coins, mainly consisting of silver.This in contrast to those issued from AD 195 onwards, which contained less than 50 % silver.It has been claimed that the "good" silver coins of the first two centuries of our era must have disappeared early in the 3 rd century within the Roman Empire: "… später hatten sie im Geldverkehr des Imperiums keine Überlebenschance."(Chantraine 1985, 412;cf. Martin 2004, 243).Implied here, as we understand it, is the presupposition that the old coins came to be regarded as bullion rather than legal tender and treated as such, i.e., melted down or exported to somewhere outside the Empire.
The idea of an export of "good" denarii towards the close of the 2 nd century or somewhat later seems to be corroborated by the presence of large numbers of hoards of such "good" coins in those parts of Europe not forming part of the Empire, i.e., in present-day Germany (to the east of the Rhine), Poland, Ukraine and Scandinavia, many of which had a final coin struck in , just before the debasement, and with the earliest coin in AD 64 or later (cf.above regarding the reform under Nero).Hoards with thousands of 1 st /2 nd century denarii are recorded, such as that of Borochitsy in western Ukraine from 1928, with 2049 coins documented out of perhaps 6000 originally found, see table IV; Mitkowa-Szubert (2000) argues that there were at least 18 kg of denarii, corresponding to 6,000 coins, with an average weight Table IV Sample  3.00 grams.The average weight of 1,687 coins in Warsaw is 2.99 g.That people outside of the Limes accepted only "good" denarii is, however, belied by the find of Schwabhausen in Thuringia in Germany from 1997, consisting of 29 denarii from AD 193 to 235, mostly from the period AD 218-235 (Henning and Mecking 2007).Some of the hoards essentially consisting of 1 st /2 nd century denarii are late deposits.The Borochitsy hoard, for instance, does not seem to have been buried until late in the 4 th -or early in the 5 th century (Mitkowa-Szubert 2000, 152).Another hoard, from Zagórzyn in Poland, unearthed in 1926 or 1927 and quickly dispersed, is said to have encompassed 3,000 or more 1 st /2 nd century denarii as well as solidi and other gold objects of the 4 th /5 th century (Kietlińska 1957, 288-290;Bursche 2000, 125-127;Bursche 2003), see table IV.It might have been buried about the same time as Childeric himself, or even later, as in the case of Smørengegård hoard from Bornholm; some 500 denarii from the 1 st /2 nd century found together with one solidus of Anthemius (AD 467-472) in the 1980s (Kromann and Watt 1984;Lind 1988, 213, No. 196A;Lund Hansen 2001, 65;Horsnaes 2013, 167-175), see table IV.All these late deposits of denarii show signs of wear, or so it seems.From the Borochitsy hoard, Lind saw 31 specimens in Warsaw the 25 th of September 1976 (i.e.Kietlińska 1957, 273-274, Nos. 14-44), and on the same occasion two coins from the Zagórzyn find (i.e.Kietlińska 1957, 288-289, Nos. 161-162;cf. Bursche 2000, 135, Photographs Nos. 1-7).Those of the Smørengegård find are about as worn as the 1,488 extant coins of the largest hoard of 1 st /2 nd century denarii from Scandinavia, that of Sindarve (originally 1,500 coins), from Gotland, found in 1870 (Lind 1981, 53-69, No. 62;Lind 2005.), making a late deposition just as likely as in the case of the Smørengegård find.
This presence of late deposited hoards with 1 st /2 nd century denarii outside of the borders of the Empire and the fact that the burial of most of the rest cannot be dated except by latest coin (often, as mentioned, 193-195), makes it an open question exactly when the export took place.For how long after AD 200 were 1 st /2 nd century denarii available in the Roman Empire, one way or another?There is nothing in favor of the argument that "good" denarii disappeared inside the Empire as quickly as has been presumed, and here we come back to the denarii in the grave of Childeric.Were 1 st /2 nd century denarii continuously available, one way or another, in Gaul from the early 3 rd to the late 5 th century AD (cf.Martin 2009)?
In northern Gaul there are several hoards denoting a long survival, well into the 4 th century, of denarii from before AD 200.Three of them are of special interest here, i.e., that of Épiais-Rhus (France, 1979), that of Beaurains (France, 1922) and that of Betteldorf (Germany, 1911).The find of Épiais-Rhus (Mitard 1985) consisted of 91 denarii from AD 64 to 189; the denarii include a Lucius Verus RIC 463 (Mitard 1985, 19, No. 78) and 416 bronze coins from 71 to 243.The hoard, thus, cannot have been buried before 243.The find of Beaurains (Bastien and Metzger 1977) is said to have consisted of c. 700 Roman gold and silver coins, of which c. 100 denarii and at least one aureus from before AD 200, the rest being gold and silver from the late 3 rd -and early 4 th centuries, with latest coin 315.81 denarii from 69-191 are determined; the denarii include one Antoninus Pius RIC 181 (Bastien and Metzger 1977, 60, No. 92.).The hoard must have been buried in AD 315 or later.The find of Betteldorf (Steiner 1912), finally, had 33 denarii from 68 BC to AD 182, 10 argentei from AD 294 to 297 and 577 small bronze coins from c. 294 to c. 324, and must have been buried in the 320s, or later.Following Lind, the denarii of the find from Betteldorf, in combination with those of Famars (see below) makes it unnecessary, as has been done (Bastien and Metzger 1977, 207 and 215;Martin 2004, 256;2009), to presume an extra-Imperial origin for the denarii in the remarkable and much discussed find of Beaurains.Denarii were made for internal circulation, not for export.If we did not have any information about finds of 1 st /2 nd century denarii outside of the Empire, an internal origin would have been the natural conclusion.
In contrast to the assemblage in Childeric's grave, where the coin of Caracalla must have been issued after 195 (as mentioned, coins with his portrait were struck only from 196 onwards), and that of Julia Domna might have been so too (as mentioned, coins with the portrait of Julia Domna seem to have been struck from 193 to 217), these three hoards lack denarii issued after the debasement in 194-195, i.e., they consist of "good" denarii (some of the denarii in the Beaurains find are, however, plated, and one of them cast.See Bastien and Metzger 1977, 47-68.).But the three hoards agree with that of the Childeric hoard in that these "good" denarii to the greater part consisted of coins from the Antonines, i.e., the period AD 138-192.
However, hoards mixing "good" denarii with bad ones, struck in AD 195 and later, are to be found all over the Roman Empire, as shown by for instance Bolin (Bolin 1958, 351-357, Supplement, Tables 7 and 8), even in northern Gaul.A large hoard of Roman gold and silver coins was unearthed in Cologne in 1909, on the precincts of the ancient Roman city.It was dug up casually and subsequently kept a secret.It is said to have consisted of 22,500 coins, in gold and silver, going from the latter part of the reign of Claudius (51-54) to the reign of Maximinus Thrax (235-238).The coins were reputedly kept in four bronze vessels.However, only c. 4,500 coins have been in some way described.Most of these, c. 4,340 in all, are denarii (FMRD VI, No. 1004.3).The earliest denarii are said to bear the portrait of Nero, and may thus, as mentioned, theoretically, belong to the latest part of the reign of Claudius, when coins with his likeness in gold and silver were issued (known coins from this hoard in fact include a Nero aureus of 51-54), but they were most probably from Nero's own reign.This stated, one could claim that the denarii went from Nero (AD 54-68) through Maximinus Thrax (AD 235-238), with almost all emperors and empresses in between represented.Plotina, the wife of Trajan (AD 98-117), and the emperor Didius Julianus (193) are missing.The reign of the latter, however, is represented by coins of his wife and his daughter.Also, some of the wives of Elagabalus (AD 218-222) and one of the wives of Severus Alexander (AD 222-235) are missing.Although the greater part of the denarii belongs to AD 193 and later, coins issued before that year are numerous enough, making up c. 1,337 in all, and most of them from the Antonine period, as is the case with the Childeric hoard.Among the pre-193 denarii there is one Hadrian RIC 244(a), two Antoninus Pius RIC 181, and one Lucius Verus RIC 463.
More difficult to assess are the finds made during quasi-archaeological excavations in 1824-1825 of the Roman baths of the ancient Fanum Martis, today Famars, not far from Tournai but on the French side of the border.A large number of Roman coins was dug up, among them, in 1824, denarii and argentei from Augustus to Constantius I, in three bronze vessels, with 3,920, 2,658 and 3,377 coins, respectively, and again, in 1825, 4,765 and 3,480 denarii and argentei from the Republic to Constantine I, in two bronze vessels.There were also, in clay pots or loose in the earth, denarii and antoniniani of the 3 rd century (Bersu and Unverzagt 1961, 186-190;TAF II, 31-33, No. 41).This is almost all the information we have.An extra-Imperial origin for the 1 st /2 nd century denarii at Famars has been suggested (Callu 1979, 9-10), as was the case for Beaurains, but the same objection is valid here.
A less debatable evidence of long survival of 1 st /2 nd century denarii is given by a hoard more to the southeast, in the French department of Haut-Savoie, on the border of Switzerland and Italy, in the modern city of Faverges, which partly covers the ancient town of Casuaria, with a mansio on the Roman road from Turin to Geneva.It mixes denarii and antoniniani.The hoard was found in the course of archaeological excavations in 1971 in the remains of a burntdown Roman building, where a bronze vessel was dug up, with content similar to that of the find of Cologne 1909.But in this case the coins were taken care of immediately.They were one drachm of Hadrian, issued in Amisus in Asia Minor (AD 130-138), 1,780 denarii from Nero to Gordian III (AD 64-241), with almost all emperors and empresses in between represented, and 525 antoniniani from Caracalla to Trebonianus Gallus (AD 215-253) (Pflaum and Huvelin 1981; cf.TAF V/2, 93, No. 17.).The hoard must have been buried in AD 253 or later.The denarii of the years AD 64-192, 783 in all, have a clear preponderance for the Antonine period, as the Tournai find.Among them, there are five issues struck for Antoninus Pius RIC 181 and one for Lucius Verus RIC 463 (Pflaum and Huvelin 1981, 47, Nos. 270-274 and 50, No. 472, respectively).
In short, official withdrawal of coins in the Roman Empire, if practiced, may have been inefficient, and people there may have been less willing to have coins melted down than has sometimes been presumed, although some melting down did actually occur (Johns 1997).It is possible to claim that denarii of the first and second centuries may have been at hand in Gaul in no small numbers as late as the 250s, and to some extent even some sixty or seventy years later, in the early 4 th century.Of still later occurrences, however, there is no real proof.It is thus unlikely that the denarii in the grave of Childeric were taken from some pool of coins in northern Gaul by the time of the death of this Merovingian ruler.Admitted, Mommsen had the idea that 1 st /2 nd century denarii circulated in what he called freies Germanien continuously from c. 200 to c. 500. 15Mommsen thus believed that the denarii in Childeric's grave proved this, because the realm of that king had become part of this same freies Germanien after the fall of the Empire in the west (cf.Mommsen 1860, 819-820).This, however, is rather speculative and an opinion not held by many today.It is also possible to argue that the denarii formed an age-long heirloom in Childeric's family, acquired sometimes in the 3 rd or early 4 th century AD, but this is unlikely, because Childeric's family does not seem to have been of long standing (Dierkens and Périn 2003).Already Merovech, the alleged father of Childeric, is half mythical.
Thus, whereas an extra-Imperial origin is an unnecessary presumption in the case of the 1 st /2 nd century denarii in the 4 th century finds of Beaurains and Famars, it must be admitted that the denarii in Childeric's grave most probably had been brought there from some place not part of the Roman Empire within the borders of c. AD 400.This brings us back to the odd coins in the grave, the Republican denarius from before 30 BC and the 4 th century AD siliqua of Constantius II.To take the latter first, no hoard combining 1 st /2 nd century denarii with silver coins of the latter half of the 4 th century is known from Gaul or elsewhere on Roman territory as it was in the 4 th century, but there are some from outside the borders.There is the 19 th century find of Lengerich in Lower Saxony in Germany, consisting of 2 nd century denarii, gold objects, siliquae and solidi of the 4 th century (FMRD VII, Nos. 1033-35;Martin 2009).There is also an early 20 th century find combining denarii and siliquae from the non-Roman part of Hungary, Kecel I (Alföldi 1920(Alföldi -1922;;Alföldi 1923), to distinguish it from Kecel II, a large hoard of denarii found in 1934 (Jónás 1935;Biro-Sey 1987;Bland 1997;Stribrny 2003, 18-30).More recently a mixed hoard of denarii and siliquae -yet to be published -is said to have been unearthed in Gudme on Funen in Denmark (Horsnaes 2010, 139).Gudme also boasts the largest hoard of 4 th century siliquae found in the non-Roman parts of Europe, 293 coins from AD 337 to 366, recovered in the 1980s (Kromann 1988;Horsnaes 2010, 138-139).
The most important and least debatable find in this category is the one found in Laatzen in Lower Saxony in 1967, with 74 denarii (including imitations) from Vespasian to Commodus  and four 4 th century silver coins, i.e., two of Constantius II, one of Julian and one imitation of Constantius II (Zedelius 1974), see table IV.
Obviously there was an outflow of 4 th century silver coins from the Empire, less distinct and of smaller proportions than that of the 1 st /2 nd century denarii (for the hoards of Zamość (Poland) and Orgeyev (Moldova), see Kropotkin 1970).Given the fact, that 1 st /2 nd century denarii were still at large in non-Roman Europe in the 4 th and 5 th centuries, shown for instance by the above mentioned hoards of Borochitsy, Zagórzyn and Smørengegård, it is natural to presume that these late silver coins sometimes were added to such assemblages of older silver coins.
As to the Republican denarius, it is of relevance that large hoards of 1 st /2 nd century silver coins from outside the Roman Empire at times contain one or two denarii of the Republic, notably the abovementioned Kecel II hoard, originally almost 3,000 coins (as for original size, see Jónás 1935; as for Republican denarii, see Biro-Sey 1987, 31-32), and the hoard from Drzewicz in Poland, with perhaps 1,600 coins (Krzyżanowska 1976, 75).Both hoards have their respective final coins dating to the second decade of the 3 rd century.
It has been argued that there might have been a connection between the Laatzen find and the two hoards from Kecel, seen as one assemblage (Zedelius 1974, 43-44 and 53).The idea that the two Kecel finds should be two parts of one single hoard must be regarded as somewhat hypothetical, but a real connection exists between the Laatzen find and Kecel II, in the form of two ancient imitations of denarii, die-linked to each other, one in the Laatzen find, the other in Kecel II.There is also a connection of the same kind between a hoard of some 250 denarii found on Gotland and the Kecel II find (see Stribrny 2003, 59 and 58, Map), all this implying secondary movements of Roman coins outside the Limes over long distances.The geographical distribution of these imitations (Stribrny 2003;Lind 2007) makes a (nonofficial) Roman origin unlikely.It thus is difficult to claim that the numerous hoards of 1 st /2 nd century denarii found outside the Empire (see Fig 10) essentially constitute working-material for silversmiths.The hoards must have been kept for symbolical as well as material value -why else bother to make new coins with portraits of the Roman emperors?
If the two Kecel finds really were one single hoard in antiquity, see table IV, it must be regarded as a hoard essentially consisting of denarii from the first and second centuries (imitations included), with a few Republican denarii, a few 3 rd century ones and a number of 4 th century silver coins added, a kind of assembly from which the 42 silver coins in Childeric's grave may hypothetically form a detached part.This is not to say that the hoard of silver coin in the grave had its origin in Hungary, only that there is nothing strange or unexpected with its composition, in a 4 th /5 th century extra-Roman context.
Finally, although there are 4 th /5 th century hoards in Barbaricum consisting of 1 st /2 nd century denarii and 4 th /5 th century gold coins -some have been mentioned, i.e.Lengerich, Zagórzyn and Smørengegård, but there are others, for instance on Gotland, see table IV -none of them have a combination of gold coins similar to that of those in the grave of Childeric.

Conclusion
We conclude that the solidus hoard together with the silver coins constituted a meaningful composition that had been manipulated for ideological purposes by Childeric's successor, Clovis.Our research shows that the precious metal objects described by Chiflet derive from at least five different sources.There are at least two different solidus hoards that have been put together and cleansed of uncomfortable elements, that is, illegitimate 5 th century western emperors.Then there are silver coins that derive from at least one denarius hoard, and a siliqua hoard, proof of past Roman grandeur and connections to leading affinities in Barbaricum.In addition, there is an array of genuine or reinterpreted Roman regalia, notably the crystal ball, the signet ring, the fibula, and the weapons.These were all signs of Childeric's past political legitimacy, and Clovis sought to arrange them in a most favorable light.Their inclusion in the burial rite meant that all Childeric's credentials were transferred onto his rightful successor Clovis.In many ways, the extravagant burial in Tournai marked one of the most important transition points on the long path from Late Antiquity to the Medieval Period known to us.