The beginning of a new page in the original publication and the pages on which illustrations were shown are indicated in blue. This facilitates quoting from the article.

page 55>>

Coins and continuity in the Dutch River area at the end of the third century AD

 

European Journal of Archaeology 6(1) 2003, 55-87.

Antony Kropff and Jos P.A. van der Vin

Abstract:  The coin series from sites in the Dutch River area show a break during the last three decades of the third century and the first decade of the fourth century AD. Coins minted for Aurelian and his successors to the throne up to Constantine I are very scarce for all sites. The break has been interpreted to indicate the end of occupation of castella and settlements around AD 275. When the site finds from the Dutch River area are presented in the form of an adapted histogram however, the coin series show a striking similarity to site finds from Roman Britain, where on the whole continuity was safeguarded during the third century. The article argues that this gap in the coin series – detectable all over the western part of the Roman Empire – is caused by the special character of coin circulation during this period in the west and does not indicate the end of activities on the site that provided the coins. Coin finds even seem to suggest continuity during this period for a number of sites in the Dutch River area.

Key words: Roman coins, numismatic method, lower Rhine limes, third-century continuity and crisis, Dutch River area.

 

Introduction

 

When we try to trace the history and development of a settlement or castellum, coins found at the site can be a very useful tool. In some cases however, coins from site finds can readily be misinterpreted.

In quite a number of site finds from castella and settlements in the west, a considerable caesura divides the coins of the Gallic usurpers that governed this area up to 274 AD from the coins of Constantine I (306-337 AD). This caesura in the coin series has often been interpreted in the light of the inroads of the Franks during the period (Van Heesch 1998:147-149; Tymann 1996:48; Van Es 1981:47). The break would indicate the end of occupation of a castellum or settlement in the mid-seventies of the third century, followed by a renewed occupation during the reign of Constantine.

But does a gap in the coin series in fact equal the end of activities on the site that

page 56>>

provided the coins? This survey argues it does not, at least not at the end of the third century in the west.

The current hypothesis, linking the caesura in the coin series to the abandonment of a settlement or castellum, is put to the test. The present survey sets out to show that this hypothesis is untenable. What is more, the numismatic data can even give support to a hypothesis of continued use of quite a number of settlements and castella in the Dutch River area throughout the third century, in spite of the gap in the coin series.

 

Theme

 

What exactly happened on the lower Rhine limes during the last half of the third century? Contemporary literary sources are very scarce, with one exception. A devastating attack by the Franks was recorded for 275 AD, but the main thrust was aimed at a more southern segment of the limes, between Trier and Strasbourg (Künzl and Künzl 1995:43; Cüppers 1990:125; Ziegler 1983:84). It remains to be seen whether the inroads of the year 275 brought devastation and a (temporary) end to Roman administration on the lower Rhine. It has often been taken for granted that it did, but looking at the numismatic evidence, a less violent scenario emerges.

Dating finds from this period proves to be a problem. In the west, a lot of ceramics and other artefacts from the late third and the early fourth century cannot be dated exactly (Willems 1989:36; Curnow 1988:61).

Roman coins from this period should not provide any dating problems, at least for as far as the moment of production is concerned. For that reason the numismatic evidence is important when questions regarding continuity of occupation and the abandonment of a settlement are dealt with.

The coins from this period (roughly 270-310 AD) however, present a number of problems that should be taken into account. Coin circulation in the western provinces differed markedly from the circulation in the other parts of the empire. 'Western provinces' should, in this context, be understood to indicate Britain, the parts of The Netherlands and Germany that once belonged to the Roman Empire, Belgium, Luxembourg, France and parts of Spain. During the period 260-274 AD, these areas were part of the 'Gallo-Roman' Empire, that broke away from the Roman Empire under the usurper Postumus. Coin circulation in this part of the empire, even long after the collapse of the separate 'Gallo-Roman' Empire, was characterized by a number of features that were lacking in other parts of the empire and which will be discussed below.

The caesura after 274 AD that can be recognized in most site finds from the Dutch River area forms the subject of this study. At first sight, the 'gap' in the coin series of a castellum or settlement seems to indicate a (temporary) end of occupation. This would imply a political-military cause of the caesura in the coin series. This can be illustrated by referring to Figure 10, a diagram (histogram) presenting the coins found at Maurik. Details of the diagram will not be dealt with for now, as they will be presented later. The height of the columns indicates the

page 57>>

number of coins for any given period, the period numbers on the x-axes refer to (sometimes clustered) periods of imperial government.

Period 18 covers coins from the period the separate Gallo-Roman Empire existed, the years 260-274 AD. Period 19 covers the years 273-296 AD, period 20 the years 296-317 and period 21 corresponds with the period 317-330 AD. At first sight it seems obvious that occupation of Maurik must have ended after 273 AD, with a slow recovery taking place after 317 AD.

However, the fact that British site finds show the same paucity of coins minted between 274 AD and the beginning of the fourth century is enough to make us stop and think again, because in Britain no break in continuity took place during this period. An illustration can be seen in Figure 4, the diagram of the coins found at Corbridge. For as far as periods 18-22 are concerned, there is a striking similarity with the Maurik diagram. Upon closer examination, a political / military cause for the caesura in the coin list of Maurik (and many other sites on the Lower Rhine) seems to be less probable than has often been supposed.

 

Coin hoards

 

In order to reconstruct the coin circulation of this period, coin hoards will be considered first. Starting from the hoards, site finds will be analyzed, followed by an interpretation of the caesura in the coin series in the west.

First of all, the way a hoard can be dated deserves attention. On the basis of dated and consecutively ordered hoards it will be possible to study the development in time of site finds. When determining the date of burial of a coin hoard found in the former Gallo-Roman Empire caution is called for. The coins minted by all emperors ruling after the coin reform carried out in 274 AD by Aurelian, up to and including the members of the Tetrarchy, are very scarce in the west. The reasons for this scarcity will not be discussed in detail here. Factors to be considered are the nature of the coin circulation in the west at the time of the reform and the economical, political and social 'climate' in the former Gallic Empire (Cheesman 1997:177; Lallemand and Thirion 1970:15). In fact these coins are so scarce in the total mass of surviving coinage in the west, that on statistical grounds alone they may well be lacking in even a large hoard. Hoards that do not contain any of these coins, and close with a coin of the Tetrici (the last of the Gallic usurpers), can therefore have been buried quite some time after 274 AD, even if a post-274 AD coin is lacking (King 1992:127; Ziegler 1983:4; Lallemand and Thirion 1970:17).

So in order to decide on the moment these hoards were concealed, an 'indirect' path must be followed. First of all, an analysis is made of all hoards (limited as the total number may be in the west) that do close with a coin struck for Aurelian, Probus and other emperors of the period between 274 AD and roughly 310 AD. These hoards can be arranged chronologically and the series then shows the changes in the composition of hoards through time. The share of coins struck in a certain period will be seen to diminish, the share of other coins will increase as time goes by. These hoards with a post-274 closing coin will give an overall picture

page 58>>

of the changes in the circulation through time, and provide a measure against which hoards without a coin of a post-274 emperor can be judged and dated with some accuracy. Hoards, all closing with a coin from 274 AD, prove to have divergent moments of burial, even up to ten years apart.

When a hoard from this area and period is considered, close attention should be given to the share of  'local copies'. The sequence of dated hoards shows these copies to provide an important dating tool.

At this point, this type of coinage needs to be introduced. Local copies can be distinguished from the regular issues on the basis of style, workmanship, size, weight and obverse or reverse type. Hoards show a peak in production of the copies between 274 and 282 AD (Davies 1992:212; Lafaurie 1975:897).

At present, the most current hypothesis states that local copies are 'Notgelt', emergency coinage, and the local producers of copies supplied the coins the official mint could or would not provide, which resulted in a shortage (Reece 2002:48, 56; Kropff 1987:13; Ziegler 1983:75; Lafaurie 1975:895).

The measure provided by the dated hoards can be taken up once more. Referring to these dated hoards, the role of the copies as a dating tool can be shown. A hoard closing with a coin of the Tetrici and containing a large share of the local copies will not be buried until some moment during the reign of Probus. A hoard containing a large share of the regular, official emissions minted for the Tetrici and only a relatively small share of the local copies will have been buried relatively early, around 274 to 276 AD (Lallemand and Thirion 1970:18). The share of local copies in this way indicates the date of burial for a hoard. Of course, analyses of the composition of the hoards will show a number of features, besides the share of local copies. These features will be dealt with in the following section.

 

Coin circulation

 

A broad outline of changes in circulation in the west can now be presented. After 'translating' the data obtained from the hoards into circulation features in this way, site finds will be dealt with.

 

The coins of Gallienus and Claudius II

The separate Gallic Empire came into being during the summer of the year 260, and lasted until the early spring of 274 AD. At the moment of Postumus’ coup, no fundamental differences existed between the coin circulation in Gaul and, for instance, the Balkans or Northern Italy. Around 260 AD the coins of Gallienus were represented in equal shares in the coin circulation of these parts. Starting from that year, coins were struck for the Gallic usurper Postumus. Halfway through this ruler's reign (260- 269 AD), Postumus’ coins began to dominate circulation in the west (Besly and Bland 1983:16). The hoards show that relatively few coins minted for Gallienus and Claudius II (produced before 268, c.q. 270 AD) penetrated coin circulation in the Gallic Empire before 274 AD (Schulzki 1989:51; Mattingly 1951:282-283). Large scale influx started only after the end of the Gallic Empire in that

page 59>>

year. The coins of Gallienus and Claudius II reached their zenith in the circulation of the former Gallic Empire in the second half of the seventies and in the eighties of the third century (Van Heesch 1998:139) and dominated circulation in that area up to 294 AD (Schulzki 1989:51).

  

The coins of the Gallic Empire

The first issues of Postumus showed a relatively high silver content, and for that reason these coins were soon driven out of circulation by the debased coinage that followed (Callu 1969:276).

The coins minted for Tetricus I (whose rule started in November 271) and Tetricus II (co-ruler with his father from the beginning of 273 on) formed the most important component of Gallic circulation. An enormous number of late third-century hoards in the west close with a coin of one of the Tetrici. These coins and their local copies dominated circulation in the west for a long period (Schulzki 1989:60).

The share of these coins in circulation can be plotted by reviewing the limited number of hoards from the west which close with a coin minted for Probus or one of his successors. The hoards from the area of the former Gallic Empire which close with a coin of Probus (276-282 AD) show a large share of the coins of the Tetrici. Hoards which close with a coin of Carus (282-283 AD) show a change. The share of the coins of Gallienus and Claudius II rose, while the percentage of Tetrici coins decreased markedly.

 

Coins minted for Aurelian and successors

As has been pointed out before, regular coins struck after the coin reform of 274 AD did not penetrate the coin circulation of the former Gallic Empire. It has been suggested these coins did circulate in the west, but with a considerable delay (Hollard 1996:208; Lallemand and Thirion 1970:15; Callu 1969:354). An analysis of hoards however clearly shows that the regular emissions struck between 274 and 310 AD never formed part of the circulation in the west, while they did in other parts of the empire (Reece 2002:49; Schulzki 1989:43, 62; Gentilhomme 1942:37). The histogram for Carnuntum (Noricum, Danube, Fig. 1) shows a considerable loss of these coins, which belong to period 19 (Schulzki 1989: 146).

For many decades the coin circulation in the former Gallic Empire consisted of the coinage of Gallienus, Claudius II, the Gallic usurpers and the local copies of these coins (Schulzki 1989:108; Thirion 1967:20). At first sight, we find a circulation that seems to be frozen in time, a static coin pool into which no coins from other parts of the empire penetrated. Further analysis however has shown that we find many changes in the content of circulation. These changes will be used in the analysis of the site in the section 'Histograms of the site finds', later in this article.

 Fig. 1 Carnuntum

Figure 1. Carnuntum - Noricum, Danube (page 60)

 

page 60>>

an analysis of site finds: theoretical background

 

An adapted histogram method was proposed by Casey and Reece (Casey 1988:41, 45; Casey 1986:88-90; Reece 1979:175-176). The number of coins for every rule or cluster of rules is divided by the number of years covered by that period. The outcome is multiplied by a factor of 1000 and then divided by the total number of coins the site yielded. This correction standardizes all sites to an equal size, by computing the yearly loss to a thousand coins. The calculated figures are presented in a bar chart, the histogram.

By using this method objective and standardized data are obtained, and a number of sites can be compared. However, not too much value should be attached to relatively small differences. Conclusions should be drawn on the basis of really striking formations in a histogram and preferably on the basis of marked differences within a number of histograms (Casey 1986:102). A drawback that cannot be obviated is the fact that no provision can be made for the time lapse between the moment the coins were produced and the moment they were lost or buried. In other words, it must not be taken for granted that coins produced 'within' a bar in the diagram circulated only during the corresponding period.  As pointed out above, coins that were produced around 274 AD could have been lost or buried at the beginning of the fourth century. This fact must be taken into account during the analyses of the histograms.

We must take care not to interpret the number of coins for a given period as the sole measure of activities on the site. In some cases an explanation can be found in the coins themselves, or in factors inherent in coin circulation (Casey et al. 1993:122).

page 61>>

Patterns that have been identified in histograms of sites all over the empire are a steady rise of the number of coins until the reign of Trajan, followed by a slight falling off under Hadrian. After that we see a marked descent under Commodus, followed by an increase after Severus Alexander, which has been linked to inflation. In the west, this development culminates in the peak for the coins of the Gallic usurpers, reflecting the near-collapse of the monetary system and the effects of hyperinflation. The peak must never be interpreted as a sign of prosperity, as it has been in the past (Casey 1988:46).

 

An analysis of site finds: looking at the west

 

Introduction

In view of this study, histograms were drawn for a number of castella and settlements in the Dutch River area. The data on which the histograms are based were calculated following the method used by Casey and Reece. The most recent publication about the site find was used, supplemented by the coins registered in the find register of the Koninklijk Penningkabinet (Royal Coin Cabinet), the National Museum of Coins and Medals at Leiden (The Netherlands) between the date of publication and 1 January 2000. Quite often a (recent) publication is lacking, and the Penningkabinet provided all data. Sites were included when the coins indicate activities on the spot around 270 AD or later.

As the coins from a number of sites from the Dutch river area have not been published before, or are dealt with in a publication not readily available to the reader, the coin data for these sites are presented in Table 1.

(Table 1. page 62)

 Table 1. Dutch River area histogram data

 

To present the Dutch histograms in a wider context, a number of sites outside the Netherlands were included. The histograms for Cologne, Flerzheim and Dalheim will be discussed at the end of this section .

 

The periodization proposed by Casey (1988:45) was used and is summarized in Table 2.

 

 Table 2 (page 63)

No.

Period

Beginning-end

No. of years

1

Augustan (Augustus, Tiberius, Caligula) 

27 B.C.-41 AD

68

2

Claudian (Claudius, Nero, Civil war)

41-68

27

3

Flavian I (Vespasian, Titus)

68-81

13

4

Flavian II (Domitian)

81-96

15

5

Trajanic

96-117

21

6

Hadrianic

117-138

21

7

Antoninian I (Ant. Pius)

138-161

23

8

Antoninian II (M.Aurelius)

161-180

19

9

Antoninian III (Commodus)

180-192

12

10

Severan I  (Septimius Severus, Geta, Caracalla)

192-217

25

11

Severan II (Elagabalus)

217-222

5

12

Severan III (Severus Alexander)

222-235

13

13

Post-Severan I (Maximinus I)

235-238

3

14

Post-Severan II (Gordian III)

238-244

6

15

Post-Severan III (Philip I)

244-249

5

16

Post-Severan IV (Decius, Gallus)

249-253

4

17

Post-SeveranV (Valerian I, Gallienus joint reign)

253-260

7

18

Gallic Empire (Postumus, Victorinus, Tetrici, Gallienus, Claudius II)

260-273

13

19

Pannonian-Illyric (Aurelian - Diocletian)

273-296

23

20

Tetrarchic (Diocletian, Maximian, Constantius, Galerius, Constantine I)

296-317

21

21

Constantinian I (Constantine I, Licinius)

317-330

13

22

Constantinian II (Constantine I, Constantine II, Constans, Constantius II)

330-348

18

23

Constantinian III (Constantius II, Magnentius, Julian)

348-364

16

24

Valentinian (Valentinianus I, Valens, Gratianus)

364-378

14

25

Theodosian I (Gratian, Theodosius I, Magnus Maximus)

378-388

10

26

Theodosian II (Theodosius I, Honorius, Arcadius)

388-402

14

 

Casey and Reece in their analysis of British histograms used period 20 to cover Carausius and Allectus, the British usurpers (287-296 AD). Since these coins are lacking on most continental sites, any coins of the British usurpers included in the site finds are left out in all histograms in order to facilitate comparison of British sites to continental sites.

 

Histograms of the site finds

Introduction

At first sight, it would seem obvious to begin with the histograms of the Dutch sites. However, another point of departure was chosen. It has to be decided whether patterns in a histogram are the result of specific local causes, such as political or military ones, or were caused by events which influenced an entire area. To decide this matter, a number of sites from another area are taken as a point of reference: Britain. The hoards show that from 260 AD until the end of the third century, coin circulation in Britain closely resembled circulation in the other parts of the

page 63>>

Gallic Empire. In the third century Britain did not witness the collapse of the defence system that occurred in some areas on the continent, such as the middle Rhine. For that reason the histograms from the British sites are analyzed first.

 

The British site finds

Three sites from Roman Britain are presented: Segontium, Fig.2 (Casey et al. 1993:133), Housesteads, Fig. 3 (Casey 1988:51; Wilkes 1961:279-319) and Corbridge, Fig. 4 (Casey 1988:44).

 

 

Histo Segontium
Figure 2. Segontium, 1975-1979 excavations (page 64)

 

fig 2a Segontium (copies in period 19)

Figure 2a. Segontium (copies in period 19). 1975-1979 excavations (page 64)

fig. 3 Housesteads

Figure 3. Housesteads (page 65)

 

fig. 4 Corbridge

Figure 4. Corbridge (page 65)

 

 

A distinct peak for period 18 (Gallic Empire, 260-273 AD) is followed by a deep slump (or even a total absence of coins) for period 19 (Pannonian-Illyric emperors, 273-296 AD). This pattern is followed by an unspectacular, but noticeable increase for period 20 (296- 317 AD) and 21 (317-330 AD). The index in the histograms continues to increase even more markedly for period 22 (330-348 AD).

The period from 273 to 317 AD is represented by a relatively small number of coins struck during this period. Does this imply that the British sites that are referred

page 64>>

to in the histograms were subject to a break in the continuity of occupation? This conclusion would not be justified, because Britain did not have to deal with large-scale invasions during this period, as was the case in other parts of the empire, like Gaul. Roman Britain had to deal with rather small-scale inroads and local disturbances, but continuity as such was not threatened at this time.

page 65>>

Where political or military influences on coin circulation can be ruled out as the main factor, other influences should be taken into consideration. These influences could well have been inherent to the nature of coin circulation in the period itself. As has been pointed out, following the fall of the Gallic Empire, only a very small

page 66>>

number of coins struck in other parts of the Roman Empire after 274 came to the west. As a consequence, coins struck during period 19 are scarce or even lacking in the histograms. However, it has to be taken into account that the histograms present the coin losses related to the production period of the coins, which of course does not have to coincide with the circulation period (Casey et al. 1993:126). If it were possible (for obvious reasons, it is not) to draw up a histogram that would place the coins in their period of circulation, or strictly speaking in the period the coins were lost, then the cluster of the periods 18 through 22 would show a more even, or 'flatter' pattern. As was pointed out earlier in the section 'The coins of Gallienus and Claudius II',  the coins of Gallienus and Claudius II, produced before (some, long before) 270 AD, began to enter the area of the former Gallic Empire on a large scale only after 274 AD, circulated there in mass until the end of the third century and of course in this way also 'filled' the gap for period 19. Furthermore, as was demonstrated earlier in the section 'Coin Hoards', the peak of production of the copies lies between 274 and 282 AD, in the first part of period 19 (Reece 2002:48, 56; Casey 1993:132). In fact, this is an argument to shift all radiate copies from period 18 to period 19. The effects of this shift in the histogram are demonstrated by Figure 2a: part of the gap for period 19 that is seen in Figure 2 is now filled. This procedure will give similar results for most British sites. However, for most of the Dutch river area sites discussed here, the period 19 gap cannot be closed by shifting the copies, simply because most sites in the area have not yielded copies, or only a small number of them. This problem will be discussed in more detail later, in the section 'Will copies close the gap?'.

 

Katwijk

We cannot ascertain how long the Katwijk castellum was occupied. Plans, drawn when the remains could still be seen during the sixteenth and seventeenth century, seem to show the foundation of a late Roman horreum inside the walls, which would indicate a defensible depot dating from the fourth century. However, late Roman finds are lacking (Bechert and Willems 1995:96-97).

Katwijk yielded 62 coins. The histogram (Fig. 5) shows a clear hiatus for the period between 235 and 260 AD, followed by a peak for period 18. After that we see a marked decrease, followed by a gradual increase through period 21, the first Constantinian period.

 

fig. 5 Katwijk

Figure 5. Katwijk (page 67)

 

Arentsburg / Forum Hadriani

On the basis of the coins, the end of Arentsburg / Forum Hadriani was proposed as being around 250 AD (Jongkees 1950: 25), which is clearly too early. Later, the end of the settlement was shifted to around 270 AD (Van Es 1981:137). For this site, 287 coins were taken into account. The histogram (Fig. 6) shows a decrease after period 13, post-Severan I. The 'peak-gap-peak' pattern for period 18 through 21, which was noticeable in Katwijk and very clear in the British histograms, does not appear. No coins minted for the Gallic usurpers other than Postumus are found.

 

fig. 6 Arentsburg (Forum Hadriani)

Figure 6. Arentsburg (Forum Hadriani) (page 67)

page 67>>

Aardenburg

An end around 273 AD (Boersma 1967:76) or 276 AD (Trimpe Burger 1997:41) has been suggested for Aardenburg. The abandonment of the site has been related to inroads or invasions. It can be argued however that Aardenburg was dismantled

page 68>>

by the occupants, as the city walls were pulled down outwards (Trimpe Burger 1992:3). A natural cause for abandonment could be considered: the Dunkirk II transgression, the rising sea level in this period, threatened to flood Aardenburg.

 

fig. 7 Aardenburg

Figure 7. Aardenburg (page 68)

 

Aardenburg (Fig.7) yielded a site find of 282 coins. A very marked peak for the coins of the Gallic Empire can be noted (period 18). No increase in coin loss can be seen after this period. At first sight, the coins do not suggest continuity after the end of the Gallic Empire in 274 AD. However, a closer analysis is called for. Most of the coins from period 18, apart from ten coins for Gallienus and Claudius II, are minted for the Gallic rulers. No less than 21% of the coins from period 18 consist of local copies. This indicates, as was noted before, a terminus later than 274 AD. But how many years later? When dealing with this question, we will take two hoards from Aardenburg into account. As is common practice, the coins from these hoards were not included in the histogram of the site finds. Site finds will be thoroughly distorted when hoards are included in the analysis of stray coins.

We shall discuss hoard A, consisting of 146 coins, and hoard B, consisting of 52 coins (Boersma 1967:78-79). Both hoards close with a coin of the Tetrici and contain respectively 89% (hoard A) and 73% (hoard B) coins from period 18. In hoard B 20% of the Tetrici coins are found to be local copies, in hoard A 45%. In hoards buried around 274 AD, the share of copies of coins for the Tetrici hardly ever exceeds 7% (Lallemand and Thirion 1970:54-55). This indicates that the two Aardenburg hoards must be later: a relatively high share of local copies corresponds with a relatively late moment of burial (Ziegler 1983:74; Lallemand and Thirion 1970:55-57).

Also, hoards buried before the death of

page 69>>

Probus (282 AD) as a rule contain at least 45% of coins minted for the Tetrici, while those that were buried later hardly ever contain more than 20% of these coins (Lallemand and Thirion 1970:55; Callu 1969:349-352). As hoard A contains 70% and hoard B 48% of coins minted for the Tetrici, both hoards must have been buried before 282 AD (terminus ante quem). Hoard B can be compared to the Vught hoard, in which 22% of the coins for the Tetrici belong to the class of local copies (Kropff 1987:15). Hoard B will have been buried around the same time as the Vught hoard, for which a time of burial of 276-280 has been proposed (Kropff 1987: 16). Hoard A will date from 278-282 AD, as in this hoard 45% of the coins for the Tetrici are local copies (Van Heesch 1998:138, 145).

What does the analysis of the hoards from Aardenburg add to the picture the histogram reveals? We already saw that a substantial part of the coin complex consists of local copies, so an end date of 274 or 275 AD is not very plausible even on the grounds of the site find alone. The composition of the hoards also indicates that although the site find as such closes with coins from period 18, the end of occupation must be situated in period 19, to be more precise in the period 278-282 AD. After that the coins do not indicate site continuity, but at least some form of regional continuity emerges. It can be noted that the adjacent Oudenburg was rebuilt in stone (phase III) at the end of the third, or the beginning of the fourth century (Lallemand 1966:117). At the site coins struck for Trajan up to and including Commodus were found, followed by a break: the next coin that follows is one of Constantine I (Lallemand 1966: 118). This gives support to the theory that Oudenburg III continues the activities of Aardenburg (Boersma 1967:76) and adds weight to the hypothesis that Aardenburg was abandoned as a result of the sea level rise, because Oudenburg is more favourably situated in this respect.

 

Vechten

Figure 8 shows the histogram for Vechten, based on 1144 coins. There is an early weighting in the diagram and the usual peak for period 18 is lacking. Also, the pattern formed by period 18 through 21 is only a weak reflection of the British histograms.

 

fig. 8 Vechten

Figure 8. Vechten (page 70)

 

Grave

The site find from Grave contains only 26 coins, so cautious interpretation is called for. Even this small number of coins presents a useful picture when plotted as a histogram (Fig. 9). As a result of the small number of coins, the pattern of the histogram as a whole is fragmented. However, the trend for period 18 through 22 shows a striking resemblance to the pattern we know from the British sites: the familiar peak/gap/recovery pattern.

 

fig. 9 Grave

Figure 9. Grave (page 70)

 

Maurik

An end date at the end of the third century has been suggested for Maurik, with renewed activity in the second quarter of the fourth century (Bechert and Willems 1995:80). The site find includes 288 coins. The histogram (Fig. 10) shows peaks in period 3 and 4. These early peaks can also be seen at Vechten (Fig. 8) and Rossum (Fig. 11).

page 71>>

Maurik has a break in period 13 through 17, followed by the characteristic peak for period 18, and the subsequent pattern we know so well by now. A remarkable recovery in the histogram can be noticed in period 22 (Constantinian II, 330-348 AD).

 

fig. 10 Maurik

Figure 10. Maurik (page 71)

page 72>>

Rossum

Continuity of some kind has been proposed for the Rossum castellum, right up to the end of the Roman era (Haalebos 1976:203-204). The rivers Maas and Waal are not far apart at this point, so the spot had a great strategic importance. In the fourth century there was a stronghold on the site, probably a burgus (Bechert and Willems 1995:72).The histogram was computed on the basis of 323 coins. The histogram (Fig. 11) shows a resemblance to Maurik: early peaks and a break for period 13 through 17. The peak for period 18 is not very pronounced, and an increase in loss takes place rather late, in period 24.

 

fig 11 Rossum

Figure 11. Rossum (page 71)

 

Cuijk

On the basis of the archaeological remains, a picture of nearly uninterrupted continuity emerges. A castellum up to the end of the first century, in the second century a temple that must have belonged to the vicus, with probably a statio at some distance. During the second and third century a crossing (ford) of the river Maas, and from about 320-330 onward we find a stone bridge and a new stronghold (Bechert and Willems 1995:72). The site find for Cuijk consists of 351 coins. The histogram (Fig. 12) shows a late weighting, with a high yearly loss starting from period 22. In that respect, Cuijk can be compared to Corbridge (Fig. 4). At Corbridge the peak for period 18 is very pronounced, as opposed to Cuijk. However: the peak/gap/gradual recovery pattern shows a remarkable similarity to the trend at Corbridge, with a hesitant increase in coin loss in periods 20 and 21, and a marked growth in period 22.

 

fig. 12 Cuijk

Figure 12. Cuijk (page 72)

 

page 73>>

Nijmegen

Up to the first century AD Nijmegen had a military and political significance that exceeded the regional level. After that, even up to the fifth century, a military function remained, but it was confined to the supervision of the middle part of the Dutch River area (Bechert and Willems 1995:65). Noviomagus was eventually given up but the population, though reduced in numbers, found a refuge at the Valkhof nearby (Willems 1990:72). The castra and the stronghold on the 'Kops Plateau' no longer existed in the third century. The coins from these particular sites were left out of the histogram. Incorporating the many thousands of early coins from the 'Kops Plateau' (Van der Vin 2002) would cause an enormous increase in the yearly loss in the early period, which would make interpretation of the patterns and trends in the histogram as a whole nearly impossible.

 

fig. 13 Nijmegen

Figure 13. Nijmegen (excluding castra and 'Kops Plateau') (page 73)

 

The histogram (Fig. 13) is based on 248 coins. As was to be expected due to the 'mixed' character of Nijmegen as a whole and the differentiated pattern of occupation, the histogram does not present a very integrated picture. However, period 18 through 22 looks very familiar: a rather pronounced peak for period 18  followed by a gap and eventually by a recovery mainly from period 21 onwards (Constantinian I, 317-330 AD).

 

Maastricht

Maastricht shows a nearly unbroken continuity from the Roman era up to the present time. In the middle of the third century the town shows a decline. The inhabited area and number of inhabitants decreased (Van Es 1981:156). Under Constantine I, probably shortly after 330 AD a castellum was built which, as is

page 74>>

shown by the results of excavations, was in continuous use from the fourth through the sixth century (Bechert and Willems 1995:111). The Maastricht histogram (Fig. 14) is based on 281 coins and consequently is presented excluding the coins from the Pandhof excavation. At the Pandhof, over a 1000 coins were found, many of which date from periods 24, 25 and 26 (364 AD onwards).

 

fig. 14 Maastricht

Figure 14. Maastricht (excluding Pandhof) (page 74)

 

page 75>>

We see a very marked peak for period 18. The share of the local copies for period 18 is high: the site find contains 28% of these copies. The high share of copies and the characteristic pattern for periods 18-22 indicates continuity in the coin circulation during the second half of the third century and during the fourth century. Figure 14a shows the effect of the shift of copies from period 18 to period 19, already mentioned in the section 'Histograms of the site finds' and to be taken up in more detail later in the section 'Will copies fill the gap?'.

 

fig. 14a Maastricht (copies in period 19)

Figure 14a. Maastricht (copies in period 19). Excluding Pandhof). (page 74)

 

The other Dutch river area sites (with the exception of Aardenburg, discussed in detail already) do not yield enough copies to warrant such a dual presentation.

 

Heerlen

The histogram for Heerlen was based on 82 coins. Figure 15 shows the characteristic pattern for periods 18-22 and a resemblance to the British histograms.

 

fig. 15 Heerlen

Figure 15. Heerlen (page 75)

 

This concludes the introduction of the histograms from the Dutch River area. To be able to analyze the Dutch histograms in a broader context, a small number of histograms from adjacent regions will be presented.

 

Flerzheim

The Flerzheim 'Strassenstation' yielded 1671 coins (Schulzki 1989), forming the basis of the histogram (Fig. 16). The trends of the histograms resemble those in the British histograms, most notably Corbridge (Fig. 4). The pattern for the periods under discussion is very characteristic.

 

fig 16 Flerzheim

Figure 16. Flerzheim (page 76)

 

Nearly 45% of the coins for period 18

page 76>>

belong to the class of local copies, an indication for continuity in coin circulation after the end of the Gallic Empire in 274 AD. Figure 16a illustrates the gap-closing effect of the shift of the copies to period 19.

 

fig. 16a Flerzheim (copies in period 19)

Figure 16a. Flerzheim (copies in period 19) (page 76)

page 77>>

Cologne

Figure 17 shows the histogram for Cologne, based on 865 coins (Nuber 1984). The resemblance to Flerzheim and the British histograms is striking, even though we see a stronger weighting on periods 22-26 at Cologne. We find the by now familiar pattern for period 18-22. Coins from period 19 are lacking nearly completely, and the 'gap' in the circulation at Cologne during the years after 274 AD would partly be filled by the local copies of the coins for the Gallic rulers, which at this site amount to nearly 36% of the total number of coins for period 18.

 

fig. 17 Cologne

Figure 17. Cologne (page 77)

 

Dalheim

Dalheim (Luxemburg, Fig. 18) is based on no less than 10,951 coins (Weiler 1972; 1977; 1983). The enormous number of coins found at Dalheim for period 18 (4110 coins, 39% of which are local copies) dwarfs the rest of the histogram. And yet the histogram fits the picture obtained from the British histograms and from Flerzheim and Cologne.

 

fig. 18. Dalheim

Figure 18. Dalheim (page 78)

Will copies fill the gap?

 

Before we review the histograms, let us turn to the copies once more. Most of the copies were produced after the end of the Gallic Empire in 274 AD. Would it not make sense to plot the copies in period 19 (273-296 AD) of the histograms? It would, and Figure 16a (Flerzheim, Bonn / Cologne area) and 2a (Segontium) show that this warrantable shift would fill part of the gap for period 19. However, both

page 78>>

sites are not representative of the former Gallic Empire as a whole. Copies circulated in all parts of the (former) Gallic Empire, but this empire did not coincide with what has been called the 'Barbarous Radiate Confederation' (Reece 2002:136), the area where these coins predominate. Northern Gaul was the heartland of these copies, but also Britain and the area of the middle Rhine belonged to this 'confederation' (King 1981:97). But the content of the pool of coins in circulation in the period 260-300 AD, consisting of coins of the Central empire (mainly Gallienus and Claudius II), of coins of the Gallic Empire and of copies, differs markedly between the areas. For instance, south of the river Loire the share of the copies is rather low. The reasons for these differences are not yet clear and the mapping of the copies has not yet been carried out in any detail (Reece 2002: 136).

The share of the copies in the circulation of the Dutch river area has yet to be studied in detail, but the sites under survey indicate that the total share will not be as high as it is in northern Gaul, even though a fair number of copies keep turning up as stray finds and as a part of hoards. Aardenburg (discussed earlier) and Maastricht (Figs 14 and 14a) yielded quite a number of copies, but the other Dutch river area sites yielded no copies or only a limited number.

 

The histograms reviewed

 

The point of departure and reference are the British histograms. All periods are represented in these histograms. Period 18 (Gallic Empire) shows a clear peak, to be related to the enormous inflation at the time. We note a gap for period 19

page 79>>

(273-296 AD), followed by a gradual increase of coin loss for the periods 20 (296-317 AD), 21 (317-330 AD) and 22 (330-348 AD).

The same pattern can be noted in the histograms of sites from the limes on the middle Rhine, in present-day Germany.

 

The Arentsburg / Forum Hadriani histogram does not indicate continuity on the spot after period 18. At Aardenburg, after the end of period 18 hardly any coins are found, which would lead to the conclusion that the occupation of Aardenburg ended around 274-275 AD. However, from the large proportion of local copies belonging to period 18, and also from the two hoards from Aardenburg mentioned earlier, it can be concluded that coins continued to be used on the site until some years later, up to around 280 AD or slightly later. It also seems clear that Aardenburg will have to be considered in the light of the developments at the adjacent Oudenburg. Oudenburg III in a way continues and takes over the activities of Aardenburg. Viewed from this angle, Aardenburg does not show any local continuity after about 280 AD, but we see a distinct regional continuity into the fourth century.

 

The histograms of the other sites from the Dutch river area, except Rossum and in a sense Vechten, show the characteristic pattern for period 18-22: a high 'peak' for period 18, a gap for period 19 and a (gradual) rise in coin loss during the following periods.

 

A few tentative conclusions may be drawn. As is historically and archaeologically documented, Roman Britain as a whole shows continuity in occupation of settlements and castella during the third century and a good part of the fourth century. The British histograms will be a reflection of this continuity. Coins minted for the 'official' emperors from period 19 (273-296 AD) are nearly completely lacking, but this lack is caused by the nature of coin circulation during the second half of the third century, and not by any major political or military changes in Roman Britain. When we take the British histograms as a reference, we find the same pattern of coin loss - and coin circulation for that matter - in a number of towns, settlements and castella in the limes area in present day Germany.

Looking at the settlements and castella in the Dutch River area we find histograms that, because of their similarity with the British histograms, do not give support to the conclusions regarding a break in continuity that were drawn from the site finds in the past. It has been inferred from the gap for period 19 in the coin series that coin circulation came to a sudden end around 270 AD in the Dutch River area and adjacent areas, and coin circulation was not restored until the Constantinian period. This, however, can not be concluded from the pattern in the histograms for the third century. On the contrary, the histograms of notably Grave, Cuijk, Maurik, Nijmegen and also Heerlen and Maastricht strongly indicate continuity of coin circulation during the last three decades of the third century and the first decades of the fourth century. It has to be decided whether this proposed continuity of coin circulation could support a general hypothesis of continuity under Roman rule in (part of) the Dutch River area. To test this, a number of themes will be discussed in the following sections: the inroads by the Franks,

page 80>>

the trade routes from the Dutch River area, and the contemporary historical sources related to this period.

 

The inroads by the Franks

 

Problems related to the continuity or discontinuity of Roman occupation and administration of the Dutch river area at the end of the third century should be discussed in relation to the inroads by the Franks during the period around 275 AD. It has generally been supposed that in this area the limes collapsed and the invaders caused extensive damage (Willems 1990:45; Van Es 1981:121; Petrikovitz 1971:178): a contemporary source mentions about seventy cities that were destroyed in Gaul alone. The inroads of the Franks have been directly linked to the break in the coin series around 275 AD (Tymann 1996:48).

On the basis of contemporary sources hardly anything can be said with any certainty about the consequences of any attacks by the Franks in the Dutch River area. It remains to be seen whether the idea of an all-devastating attack by the Franks in this area deserves some nuance.

It needs to be determined where the main point of attack was situated. The Dutch River area did not belong to the areas in the west that would yield rich spoils. From a geographical point of view, and supposing the invaders would have had the objective of obtaining a quick access to the heart of the most prosperous middle and south part of Gaul, a massive attack directed at the Dutch River area would not have been a probable option. Incursions during the reign of the Gallic ruler Postumus often followed the Cologne/Tongeren/Bavay road (Cüppers 1990:125). During the period around 274 AD the Mainz/Trier/Reims road was preferred (Cüppers 1990: 125), as is illustrated by the relatively large number of hoards from this period in Luxembourg, the south east of Belgium and the area around Mainz (Ziegler 1983:84). An accumulation of hoards can also be seen along the Cologne/Dinant/Laon and Luxembourg/Arlon/Reims roads (Van Gansbeke 1955:12).

Hoards from this period are very scarce in the Dutch River area. The large Vught hoard is an exception (Kropff 1987), and apart from that the two relatively small hoards from Aardenburg are at our disposal. The hoards in this area do not indicate large-scale devastating inroads during this period. Of course the objection could be raised that not enough coins were in circulation in the area to form hoards, but the site finds from this period do not allow this assumption. And should we accept this assumption, it will be very hard to interpret the Vught hoard. In the past, this hoard was assumed to be an incidental 'import' of coins from Gaul (Lallemand and Thirion 1970:58-59), but the similarities between the composition of the hoard and that of the site finds are too striking to maintain such a claim.

Taking into account the fact that the Dutch River area does not give quick access to the heart of Gaul and considering that this area did not show a fully developed urbanization or an accumulation of rich villas, it does not seem probable that the River area suffered from inroads of the magnitude seen in Gaul. The main attacks seem to have been aimed at the middle and upper Rhine

page 81>>

(Cüppers 1990:125; Ziegler 1983:83-84), not at the lower Rhine. Finds recovered during dredging operations in the Rhine show that the Franks and the Allemanni, returning from their raids, crossed the river primarily between Strasbourg and Mainz (Künzl and Künzl 1995:43). The recently recovered hoard from Neupotz (Künzl and Künzl 1995: 3), lost in the Rhine by a body of Franks as a result of an accident during the retreat from a raid, consists of no less then 700 kilograms of metal, bronze and silver for the most part. In Gaul a raid on a limited number of villa's and a sanctuary had sufficed, but collecting a booty of this size in the Dutch River area would have taken a lot of time and trouble, if collecting it would even have been feasible at all.

With some caution, the conclusion may be drawn that large-scale devastation and disruption as a result of inroads of the Franks around 275 AD is not to be expected on the lower Rhine. There is no compelling reason to assume a significant break in continuity as a direct result of inroads around this time.

 

Trade routes and coastal defence

 

In the second half of the third century, control of the Rhine delta was still very important in view of the trade with Britain (Willems 1989:34). A considerable share of the transit trade passed through the Rhine delta, not only the trade with Britain, but also the trade with the towns and settlements on the Atlantic coast. At the time, Britain produced a surplus of corn and other food supplies and of metals (Panegyrici Latini VIII:11, 1; Nixon and Rodgers 1994:126). The trade with Britain via the Rhine did not come to an end until the first half of the fifth century, at which time the Nijmegen fort lost its function (Willems 1990:85; 1989:38).

The Roman administration would have tried by all possible means to secure the Rhine as a trade corridor, which would have implied a policy of continuity in the region. In this context it is worth noting a radical and lasting change in the trade flows from the west at the end of the third century. The importance of trade in the North Sea and the Atlantic area increased, whilst trade oriented on the Mediterranean decreased.

Securing the trade routes therefore also implied the defence of the coasts in this region. It has been argued that Aardenburg probably formed a part of the coastal defence, especially under Postumus (Boersma 1967:76). Research has shown that the 'Saxon shore system' (Litus Saxonicum) was based in essence on a third-century initiative, and while Aardenburg formed a part of the earliest phase, Oudenburg formed a part of the later stages of the system (Drinkwater 1987:221).

Summing up, we may conclude that the delta of the Rhine, Maas and Schelde and the connected coastal area certainly cannot be regarded as a marginal part of the empire: on the contrary the rivers and coast formed an infrastructure of great importance and well worth the defence effort.

Now we turn to factors, other than of a strictly military kind, which could have influenced this infrastructure. Before we look at the contemporary historical sources on the River area in this period, we turn to the sea level changes at the time.

page 82>>

Sea level Rise: the Dunkirk II transgression

 

If the inroads by the Franks did not result in devastation and a break of continuity in the Dutch River area, it still remains to be established whether other factors might have influenced continuity in this region.

In the third century a water level rise occurred in the delta, as a consequence of a slow rise in sea level. This water level rise resulted in the loss of agricultural land. There was no question of a sudden disaster, but the archaeological remains dating from the last part of the third century show the effect on the occupation of settlements (Bloemers 1990:116). In the long run inhabitation of most of Zealand became impossible (Boersma 1967:65). Around 300 AD even the higher parts of Zealand were affected. Originally Roman Aardenburg seems not to have been limited to the walled area. It has been presumed that in the course of the third century occupation was confined by the rising water level to the walled area, which was situated on higher ground (Trimpe Burger 1997:6). It seems quite likely that the abandonment of Aardenburg was caused by the rising sea level.

The panegyric on Constantius, the subject of the following paragraph, deals in some detail with the high water level in the delta.

 

 Contemporary historical sources

 

We have only one contemporary historical source for the River area in this period: the official speech on Constantius, found in Panegyrici Latini VIII (Nixon and Rodgers 1994; Herrmann 1991). Constantius is congratulated on behalf of the city of Autun on the occasion of the recovery of Britain after the campaign against the British usurpers. The text also deals with events during the last decades of the third century, starting at about 260 AD. Present day commentators have pointed out the historical accuracy, not to be found in the work of other authors in this genre (Nixon and Rodgers 1994:105). The speech was probably delivered in the early spring of 297 at Trier, while the campaign itself took place in 296.

Two themes from the panegyric are discussed here: the nature of the soil and the water level in the River area, and the pacification of this region. The first theme is dealt with in no uncertain terms, but on the second theme the text is not very explicit. All references that follow are to the sections of Panegyrici Latini  VIII.

The author writes elaborately about the condition of the delta, and observes that it consists hardly, or even not at all, of soil in the strict sense. The soil is soaked and saturated with water. Feet sink into the ground: the surface moves and yields when trod upon. As the topsoil seems to float on water, the region seems best suited for training soldiers in the art of naval warfare (8.1-4). Reading the panegyric in detail, one gets the impression the soldiers were troubled more by the high water level than by hostilities in the River area, about which the text is not so explicit. This seems strange, for it is in the nature of the Panegyrici Latini to give a detailed description of even relatively insignificant military victories.

The author does mention however, that while preparing for the campaign against the British usurpers, the River delta was cleared from enemies (7.4, 8.1).

page 83>>

The exact nature of the operation is not clear: notwithstanding the treacherous nature of the soil and the many hiding places in the woods, we are told the barbarians were obliged to submit to Roman rule (8.4). The end result of this submission is mentioned: 'It is now for me that the Chamavians and Frisians plow' (9.3) but there was no hard gained victory, or at least the author does not mention this. The submitted tribes contributed in full to society: fields that had fallen into a state of neglect were cultivated again, cattle were brought to the market and food supplies were produced (9.3-4). When the tribes are summoned to the levy, they do not object (9.4). All this seems to allude to a rather mild kind of pacification, not to a difficult campaign to restore a collapsed border. When the author finally mentions units of Franks that were destroyed, it concerns mercenaries serving under Carausius in Britain (17.1-2; Hermann 1991:638).

 

Conclusion

 

On the face of it, the coins in site finds from settlements and castella in the Dutch River area seem to support the hypothesis that Roman authority over the region came to a sudden, temporary end around 275 AD. The coin series shows a break: hardly any coins struck for the emperors ruling after 274 are found. The break in the coin series lasts well into the Constantinian period.

A first indication however that the break in the coin material does not necessarily have to indicate an interruption of continuity is found in the British site finds. In Roman Britain, continuity during the third century is evident and yet we find the same break in the site finds.

 

Our survey shows that the break in the coin series, contrary to what has been supposed in the past, cannot be linked to military or political causes in the strict sense, such as 'barbarian' inroads, resulting in devastation or the evacuation of castella and settlements. The break in the coin series seems to be inherent to the nature of the coin circulation and can be related to economic and monetary causes. The composition of the coin hoards shows that the coin reform carried out by Aurelian in 274 AD was totally ineffective in the west. Coin circulation in the areas that had been part of the separated Gallo-Roman Empire began to deviate markedly from circulation in other parts of the empire. Circulation in the west was characterized by a number of features. Coins struck for the Gallic usurpers and the local copies that were inspired by these coins played an important role. Coins struck for the emperors Gallienus and Claudius II entered circulation in the west with a considerable delay. And, most outstanding feature of all, the coins of Aurelian and all the emperors that succeeded to the throne up to the reign of Constantine I entered circulation in the west in very small numbers.

Thus the hoards show that the break in the coin series formed by the site finds does not substantiate a break in continuity of occupation and administration in the Dutch River area, nor in any other part of the region that once formed the Gallo-Roman Empire.

Can we on the other hand 'prove' continuity on the basis of the histograms of settlements and castella? Some caution is called for. Numismatic facts can hardly

page 84>>

ever 'prove' anything outside its own field. Numismatics can only provide data, which can be used to judge whether a hypothesis outside its field is probable or not probable. However, the numismatic data produced in this survey show that if there was continuity in (part of) the Dutch River area during the third century, the histograms would present about the same picture they do now. That in itself is an important recognition. What is more, the striking similarity of late third-century patterns in histograms from the Dutch River area to histograms from sites with a proven third-century continuity strongly indicates continuity in this period for the Dutch sites also.

 

In addition to the data that can be obtained from the site finds, other facts seem to suggest continuity. First of all, the main inroads by the Franks of this period were aimed against the centre of Gaul, and the Rhine was crossed mainly between Cologne and Strasbourg. The objective was the use of the main routes into the heart of Gaul. The attacks probably did not cause much damage in the Dutch River area. Looking at the contemporary source with regard to the area at the end of the third century, the Panegyrici Latini VIII, we do not find clues indicating a hard fought recovery of the River area by the Roman army. On the contrary, a limited pacification, a consolidation of Roman administration seems to have been the objective of Roman operations. Where the author describes only the final outcome without giving details about the nature of the operation, he discusses the effects of the high water level in detail. The archaeological data do not support the hypothesis of a large-scale destruction around 275 AD: a uniform burn layer cannot be identified.

 

To conclude: the site finds of coins from a number of settlements and castella in the Dutch River area do not support the hypothesis that Roman continuity broke off abruptly around 275 AD, only to be restored in the first decades of the fourth century. On the contrary, the histograms seem to give support to the hypothesis that Roman power and presence might have been weakened, but this did not result in a break in continuity.

 

Acknowledgements

The authors wish to express their thanks to Willem J. H.Willems for his comments and contributions. They are also grateful for the comments and contributions of Mark Pearce and two anonymous referees.

 

References

BAART, J.M., 1990. Inventarisatie van Romeinse muntvondsten in Noord- en Zuid-Holland. Amersfoort (NAR 12).

BECHERT, T. and W.J.H. WILLEMS, 1995. De Romeinse rijksgrens tussen Moezel en Noordzeekust. Utrecht.

BESLEY, E. and R. BLAND, 1983. The Cunetio Treasure, Roman Coinage of the Third Century AD. London.

BLOEMERS, J.H.F, 1990. Lower Germany, military organization and its role in the study of a frontier zone. In Akten des 14. Internationalen Limeskongresses 1986 in Carnuntum: 111-120. Vienna.

BOERSMA, J.S., 1963. Romeinse muntvondsten in Noord-Brabant. Jaarboek voor Munt- en Penningkunde 50: 1-75.

 

page 85>>

BOERSMA, J.S., 1967. The roman coins from the province of Zeeland. Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek, 17: 65-99. Amersfoort.

CALLU, J.-P., 1969. La politique monétaire des Empereurs Romains de 238 à 311. Paris.

CASEY, P.J., 1986. Understanding Ancient Coins. London.

CASEY, P.J., 1988. The interpretation of Romano-British site finds. In P.J. Casey and R. Reece (eds), Coins and the Archeologist: 39-56. Second edition, London.

CASEY, P.J., J.L. DAVIES, and J. EVANS, 1993. Excavations at Segontium (Caernarfon) Roman Fort, 1976-1979. London. (Council for British Archaeology Research Report 90).

CÜPPERS, H., 1990. Die Römer in Rheinland-Pfalz. Stuttgart.

CURNOW, P., 1988. Coin lists: some problems of the smaller site. In P.J. Casey and R. Reece (eds), Coins and the Archaelogist: 57-72. Second edition, London.

DAVIES, J.A., 1992. Barbarous radiate hoards. Oxford Journal of Archaeology 11(2): 211-224.

DRINKWATER, J.F., 1987. The Gallic Empire, Separatism and Continuity in the North-Western Provinces of the Roman Empire AD 260-274. Historia 52, Stuttgart.

ES, W.A. van, 1981. De Romeinen in Nederland. Second edition. Bussum.

GANSBEKE, P. van, 1955. Les Trésors Monétaires d’Époque Romaine, RBN 101: 5-21. Brussels.

GENTILHOMME, P. le, 1942. La trouvaille de la Vineuse. Revue Numismatique 5(6):23-50. Paris.

HAALEBOS, J.K., 1976. Munten uit Maurik. Oudheidkundige Mededelingen uit het Rijksmuseum van Oudheden te Leiden  57: 197-226. Leiden.

HEESCH, J. van, 1998. De muntcirculatie tijdens de Romeinse tijd in het noordwesten van Gallia Belgica. Brussels.

HERRMANN, J., 1991. Griechische und lateinische Quellen zur Frühgeschichte Mitteleuropas bis zur Mitte des 1. Jahrtausends U.Z.. Berlin.

HOLLARD, D., 1996. La Circulation monétaire en Gaule au IIIe siècle après J.-C.. Studien zu Fundmünzen der Antike 10: 203-217. Berlin.

JONGKEES, J.H., 1950. De Romeinse muntvondsten van Vechten en Arentsburg. Jaarboek voor Munt- en Penningkunde 37: 1-26.

KING, C.E., 1981. The Circulation of Coin in the Western Provinces AD 260-295. In A.R. Hand and D.R. Walker (eds), The Roman West in the Third Century: 89-126. Oxford: British Archaeological Reports (International Series S.109).

KING, C.E., 1992. The Chalgrove Hoard. In Roger Bland (ed.), Coin Hoards from Roman Britain 9: 125-132. London.

KROPFF, A.C., 1987. A hoard of Third-century Roman coins from Vught. Jaarboek voor Munt- en Penningkunde 74: 5-20.

KÜNZL, E. and S. KÜNZL, 1995. Ein Schatz aus dem Rhein: der grosse Kiesbaggereifund von Neupotz. Rheinzabern.

LAFAURIE, J., 1975. L’Empire Gaulois, apport de la numismatique. In Aufstieg und Niedergang der römischen Welt 2: 967-1005. Berlin and New York.

LALLEMAND, J., and M. THIRION, 1970. Le trésor de Saint-Mard 1. Wetteren.

LALLEMAND, J., 1966. Monnaies Romaines Découvertes à Oudenburg. Helinium 6.

MATTINGLY, H., 1951. The Clash of the Coinages circa 270-296. In P.R. Coleman-Norton (ed.), Studies in Roman Economic and Social History in Honor of Allan Chester Johnson: 275-289. Princeton.

page 86>>

NIXON, C.E.V. and B.S. RODGERS, 1994. In  Praise of later Roman emperors: the Panegyrici Latini. Berkeley and Los Angeles.

NUBER, E., 1984. Fundmünzen der römischen Zeit in Deutschland, 7, 1, 1. Berlin.

PETRIKOVITZ, H. von, 1971. Fortifications in the North-western Roman Empire from the third to the fifth Centuries A.D. Journal of Roman Studies 61: 178-218. London.

REECE, R., 1979. Zur Auswertung und Interpretation römischer Fundmünzen aus Siedlungen. Studien zu Fundmünzen der Antike 1: 175-195. Berlin.

REECE, R., 2002. Coinage of Roman Britain. Brimscombe Port.

SCHULZKI, H.J., 1989. Die Fundmünzen der römischen Strassenstation Flerzheim. Keulen.

THIRION, M., 1967. Les Trésors Monétaires Gaulois et Romains trouvés en Belgique. Brussels.

TRIMPE BURGER, J.A., 1992. Romeins Aardenburg. Aardenburg.

TRIMPE BURGER, J.A., 1997. De Romeinen in Zeeland. Goes.

TYMANN, I.D., 1996. Romeinse Munten uit Vechten. Oudheidkundige Mededelingen uit het Rijksmuseum van Oudheden te Leiden 76: 139-152.

VIN, J.P.A. van der, 1986. Munten uit Arentsburg. Westerheem 35(5-6): 219-224.

VIN, J.P.A. van der, 2002. Die Fundmünzen der römischen Zeit in den Niederlanden 3,1. Mainz.

WEILER, R., 1972. Die Fundmünzen der römischen Zeit im Grossherzogtum Luxemburg 1. Berlin.

WEILER, R., 1977. Die Fundmünzen der römischen Zeit im Grossherzogtum Luxemburg 2. Berlin.

WEILER, R., 1983. Die Fundmünzen der römischen Zeit im Grossherzogtum Luxemburg 3. Berlin.

WILKES, J., 1961. Excavations in Housesteads Fort. Archaeologia Aeliana 4th series, 29:  279-319.

WILLEMS, W.J.H., 1989. Das Rhein-Maas-Delta als Grenzgebiet vom 3. bis zum 8. Jahrhundert. In Siedlungsforschung 7:31-49. Bonn.

WILLEMS, W.J.H., 1990. Romeins Nijmegen. Utrecht.

ZIEGLER, R., 1983. Der Schatzfund von Brauweiler. Köln and Bonn.

 

 

Biographical Note

 

Antony Kropff obtained his MA in Classical Archaeology at the University of Leiden. He has worked on several projects in the field of Roman coin hoards and site finds from the third century AD, and has co-operated before with Jos van der Vin in the publication of the Vught hoard.

page 87>>

Jos P.A. van der Vin is Curator of Ancient Coins and also reader of Ancient Numismatics at Leiden University. He has published extensively on Roman and Greek coins, and is editor with Maria R.-Alföldi of the Die Fundmünzen der Römischen Zeit in den Niederlanden series (Mainz: Akademie der Wissenschaften und der Literatur).

 

Korte samenvatting:
De muntreeksen van sites in het Nederlandse rivierengebied vertonen een breuk gedurende de laatste drie decennia van de derde eeuw en het eerste decennium van de vierde eeuw n.Chr. Munten geslagen voor Aurelianus en zijn opvolgers tot Constantijn I zijn op alle sites zeer schaars. De breuk wordt wel geïnterpreteerd  als een teken dat het gebruik of de bewoning van castella en nederzettingen rond 275 n. Chr. afbrak. Als de muntvondsten uit het Nederlandse rivierengebied echter worden gepresenteerd in de vorm van een histogram, dan vertonen de muntreeksen een opvallende overeenkomst met de muntreeksen van castella en nederzettingen uit het Romeinse Britannia, waar over het algemeen genomen gedurende de onderzochte periode de continuïteit gewaarborgd was. Het artikel beargumenteert, dat de breuk in de muntreeksen -die zich overal in het meest westelijke deel van het Romeinse rijk voordoet- zijn oorzaak vindt in het speciale karakter van de muntomloop in dat gebied gedurende deze periode. De breuk in de muntreeksen bewijst dan ook niet dat er een eind kwam aan de activiteiten op de betreffende sites. De muntreeksen lijken er eerder op te wijzen dat er in een aantal Nederlandse castella en nederzettingen sprake was van continuïteit in bewoning of gebruik, ook na 270 n.Chr.

Kernwoorden
:
Nederlands rivierengebied, Rijnlimes, numismatische methoden, Romeinse munten, crisis en continuïteit gedurende de derde eeuw n. Chr.