First published on this website May 11, 2013



Demonetization of Roman coins:
Saloninus Augustus, a case study.


In 260 AD, Gallienus’ youngest son was in residence in Cologne as Caesar while the Emperor himself set up headquarters in Milan. In the summer of that year, governor Postumus rebelled, took the purple and marched on Cologne with his troops. The city fell to him and Saloninus was killed.
The Cologne mint had produced coins for Saloninus Caesar, as other mints in the Empire have done. However, Cologne was the only mint also issuing coins for Saloninus Augustus.

We will discuss the Saloninus Augustus issue. Why and on whose authority were the coins produced?
A comprehensive die study will demonstrate the die links within the series; the implications are examined. Furthermore, the size of this coin issue, the survival rate and the occurrence of these coins in hoards will be discussed.
The Saloninus Augustus coins were thought to be seized and melted down by order of Postumus after Saloninus’ death. We will argue that this demonetization of coins did not take place. Demonetization of coins is a modern concept which should not be projected into antiquity.

1. Postumus and Saloninus: history and coins 

The main written sources (Zosimos 1.38; Zonaras 12.24.10-12) give the sequence of events. Postumus quarrelled with Saloninus and his guardian Silvanus over booty, recaptured from a raiding party of Germans after an inroad was cut off. Postumus distributed the spoils among his troops but Silvanus declared that everything should be surrendered to the state treasury. In response Postumus incited his troops to revolt, took the purple and marched on Cologne, executing both Saloninus and Silvanus after the siege had ended. The written sources are silent on the date of the siege of Cologne and the death of Saloninus, but both the last papyri mentioning the prince [1] and the last emission for Saloninus as Caesar from Alexandria date from late August 260 AD [2]. Taking into account that the news of the death of Saloninus took a couple of weeks (or even months) to reach Egypt, the events in the west can be placed between June and August of that year.
The written sources are also silent on an elevation to the rank of co-emperor with Gallienus which the Saloninus Augustus coins seem to suggest. The title of Augustus was probably adopted by Saloninus on an
ad hoc basis. After his usurpation, Postumus immediately ordered the production of coins in order to pay the troops loyal to him with coins bearing his portrait, name and the title of Augustus. To counter Postumus’ rebellion, the authority of the resident member of the imperial family was emphasized and even augmented on the coins [3].
The present consensus is that the Saloninus Augustus coins were indeed produced in Cologne [4] and not in Trier as has been thought in the past. No other mints produced coins for Saloninus Augustus. The Augustus title is not otherwise accredited to him and so was not officially recognized by Gallienus [5].

The coins of Saloninus as Caesar are common, but the coins of Saloninus as Augustus are rare: at the moment (April 2013) 59 surviving antoniniani and one aureus are known.
In 1979, Norman Shiel published the then known 15 coins [6]. Subsequently, more coins have been identified and published [7].
Shiel established die links within the series of the then known pieces. He concluded that the whole issue was produced under siege conditions. This would be evident from the limited range of types, apposite in themselves, the incidence of die linking between the coins, and the paucity of extant specimen. The great majority of the issue would have been seized and melted down upon the capitulation of the city [8].
The publications following Shiel’s contribution presented new coins but did not discuss new research issues. A few new die links were established, other proposed links were rejected [9]. No comprehensive die study was carried out and the hypothesis on the demonetization of the Saloninus Augustus coins was not tested.

2. Die duplicates

As we noted before, the high incidence of die linking between the coins was thought to be a clear sign of production under siege conditions [10]. After the first die duplicates were published by Shiel, only a few new die identities have been recognized in subsequent publications. For this contribution, we have carried out a comprehensive obverse die study of all known coins.

Die comparison requires patience and is quite laborious [11]. Coin obverses from the same issue tend to have quite a lot of similarity, as a result of the way coin dies were produced. The negative, ‘female’ coin die (matrix) was prepared with the imprint of a positive, ‘male’ punch die to transfer the image, in this case the portrait. The die cutter subsequently reworked, cut and detailed the negative coin die [12]. This procedure saved time and produced coin dies which looked very much alike.
To differentiate between coins of similar but not identical dies, the legends on the coin are helpful. A second, separate die cutter fitted the legends in the space left by the portrait cutter [13]. The position of the characters in relation to the portrait, the spacing, unusual character shapes and breaks in the legends can be vital to ascertain die duplicates [14]. 

Of all known specimen, the obverses were compared. As many coins are in private collections, this was done with the aid of photos. A database was made of all distinguishing characteristics of the coins to identify die duplicates, for instance the position of the individual rays of the radiate crown in relation to the legends; form and position of the draping of the bust and details in relation to the portrait. A few examples of coin entries can be seen here.
We will illustrate a few aspects of the procedure of die comparison. At first sight, four of five coins below seem to have come from the same obverse die [15].

GS 44

SS 45

GS 42

coin 1

coin 2

coin 3

GS 43

GS 8

coin 4

coin 5

The portrait with the rendering of the hair, the radiate crown and the drapery on the right shoulder are very similar. The coin of fig. 4 clearly deviates from the other four coins: the neck is longer and more slender [16].
All coins are in private collections. The coins of figure 1 and 2 are clearly die duplicates. The coin of figure 1 is weakly struck, but the legends give away the die duplicate: observe the form of the ‘IMP’ lettering with the typical break in the top of the P, the form of the R in Valerianus and the shape and position of the G of ‘AUG’.
Coin 3, although very similar to coin 1 and 2, is not a die duplicate, as is first of all shown by the form of the short ribbons in the back of the radiate crown: pointed in figure 1 and 2, ending in a chisel mark in figure 3. The drapery on the right shoulder is rendered in a simplified form in coin 3. Also, the position of the individual spikes of the radiate crown in relation to the lettering is different and the typical break in the top of the P is lacking.
The obverse of coin 5 seems also to be identical with coins 1 and 2. However, coin 5 is not a die duplicate of these coins, as is shown by the position of the second ray from the top of the radiate crow in relation to the L in the legend and the position of the fourth ray from the top in relation to the O, when comparing with the coin of figure 2. 

In this way, the obverses of all known 59 antoniniani were compared. Photos of all known specimen are presented in this table. In this table, the now known dies are indicated by a character and in the row behind this character all coins struck from this obverse die are presented. From some dies only one coin survived, from other dies three or four. The coins bear a number according to the system introduced by Hans Gilljam in his publications on the subject. The coin numbers are preceded by either the prefix GS (Gilljam, Spes reverse) or by GF (Gilljam, Felicitas reverse). The last coins published by Gilljam himself so far are GS 42 and GF 11. Following Sondermann, we have carried on Gilljams number system for new coins. Of the total of 59 antoniniani, ten are in a museum or collection open to the public, 19 were from twelve hoards (some of them now in public collections) and 30 are in private collections or in the coin trade. The Spes Publica reverse is represented by 47 coins, the Felicitas Augg reverse by twelve coins. The quality of six coins or photos was insufficient to allow analysis (see bottom part of table and one coin (a Felicitas Augg) is no longer traceable [17].
Of the remaining 52 coins, fifteen were produced from a die only recognized once (‘unique die’). For the production of the other 37 coins, fourteen obverse dies were used.  

Photos of two of the coins are as yet unpublished: one coin(GS 46) is in the Cambridge Museum of Archaeology and Anthropology and has been mentioned by Robertson [40] and one coin (GS 47) was in the coin trade in April 2013 [18]. A photo of one coin (GF 7) is missing as the coin was lost. 

The conclusion is that the Saloninus Augustus issue includes quite a number of die duplicates. We will use the data to test the demonetization hypothesis.  

3. Demonetization of the Saloninus Augustus coins?

 Demonetization, the withdrawal and melting down of the coins is a recurring theme in the literature dealing with the coins of Saloninus as Augustus [19]. The very small number of surviving specimen has been taken as a sign that the issue has been captured and melted down upon the fall of the city which produced the coins. To test the demonetization hypotheses we would have to determine the size of the issue and the survival rate of coins from the issue, and compare these data with other issues. 

3.1. Production of the Saloninus Augustus emission 

We now have 59 antoniniani for Saloninus Augustus at our disposal. The 52 coins which could be analyzed were produced from 29 separate obverse dies.
Two reverse types are found within the Saloninus Augustus emission: Spes Publica on antoniniani [20] and Felicitas Augg on an aureus [21] and on antoniniani [22]. In
Cologne, the Spes reverse had already been used for Saloninus Caesar [23]. The Felicitas Augg reverse had not been used for any of the Cologne emissions of the house of Gallienus before. It stands the reason that the coin production for Saloninus Augustus started with the Spes reverse to save time, as the reverse dies were already available. The Felicitas Augg dies were cut especially for this emission. This reverse is more fitting to the occasion than the youthful Spes type: the Augg added extra weight to the Saloninus’ claim that he was now co-Augustus with his father Gallienus. We may assume that the Felicitas Augg coins followed the Spes Publica coins when the Spes reverse dies ran out. The small number of Felicitas coins may indicate that this emission had just started when production was cut short by the fall of Cologne.
In fact, looking at the Spes and Felicitas coins, it is probably correct to identify these coins as the product of two distinct emissions [24]. 

3.2. The size of the emissions 

To test the demonetization hypothesis, we will try to estimate the size of the emissions and determine the survival rate for these coins.
First of all, the production for each individual die can be estimated. Modern practical minting trials have been performed, using conditions and methods closely resembling ancient procedures [25]. A minimum production for each antique obverse die has been estimated at 10.000 to 16.000 coins. Statistical analyses have yielded similar figures [26].
More recently, the production for an individual obverse die has been calculated between 10.000 to 40.000 coins [27]. A production between 10.000 and 15.000 coins for an obverse die has also been proposed [28].
Based on plausible considerations, ancient numismatics have estimated that out of the total number of coin dies in production in antiquity, we might now only know about half by their surviving coins [29]. In consequence, for the Saloninus Augustus emissions, between 29 (the number of obverse dies identified in this survey) to up to 58 obverse dies might have been used. But we will have to allow for some upward margin, as one coin is lost and six coins could not be linked to a known die, because of the poor quality of the coin or photo. Based on the number of dies known from only one coin in relation to the number of dies represented by more coins, we will assume that these seven coins could add two new dies, bringing the estimated total number of dies in production somewhere between 31 and 62.

When the city fell to Postumus, not all dies would have reached their maximal production, especially as the Felicitas Augg emission seems to have been cut short, judging by the small number of surviving specimen. For this reason we will limit the average production for an obverse die at 5.000 to 10.000 coins. The total production therefore might have been between 155.000 (31 dies, 5000 coins a die) and 620.000 coins (62 dies, 10.000 coins from a die). Total production might have been below the lower limit, if production from many dies would have been cut short by the fall of the city. Total production will however very probably not have exceeded the highest estimate of 620.000 coins.

This is a rather small, but by no means insignificant production. In 1979 the coins were thought to be extremely rare, but more recently it was concluded that the Saloninus Augustus emissions are ‘nicht ganz schwach repräsentiert’ [30]. This conclusion is confirmed by the hoard of Eauze, containing no less than six coins for Saloninus Augustus [31]. 

3.3. Demonetization, a modern concept 

Demonetization is a modern concept, based on the fact that most coins became token money during the twentieth century and by the consideration that banknotes are fiduciary currency. Both types of currency can easily be withdrawn from circulation and replaced by a present day modern central bank with a firm grip on circulation.
To project demonetization on Imperial Roman coin circulation is misleading. Was demonetization ever successfully practiced in ancient history? Even when an ancient source hints at the withdrawal of certain coins during the third century [32], the coin circulation of the time shows no sign of demonetization of these coins [33]. Barrett [34] has pointed out that official demonetization of Roman coinage was a difficult and complex process, and it is not clear how, in practical terms, it could have been carried out. After discussing the straightforward methods to demonetize current currency, Barrett states that ‘Roman coinage is of course a different matter. Its value […] resided essentially in the value of its metal content. The stamp of the issuing authority was not so much a promise to redeem but, rather, confirmation that the weight and purity of the metal conformed to a recognized standard’.

To summarize, the modern concept of demonetization should not be projected onto ancient economy without the following evaluation of the numismatic evidence. 

3.4. The survival rates of ancient coins 

We have discussed the average production to be expected for an obverse coin die. How many of the coins produced did survive to the present day? Throughout history, coin emissions can be identified for which two parameters can be determined: the approximate number of coins produced and the number of surviving coins. François de Callataÿ has published a diachronic study of a number of these emissions, five ancient, four medieval and three modern [35].
The survival rates range from 1: 2.000 to 1: 10.000 or more, while most of the values found cluster around the ratio of 1: 5.000 [36].
For the antoniniani of Saloninus Augustus, we estimated total production between 155.000 and 620.000 coins. At the end of April 2013 when this survey was completed, some 59 surviving antoniniani have been published. With a total production of antoniniani between 155.000 and 620.000 coins and 59 surviving antoniniani, the rounded off survival rate can be calculated to be between 1: 2.600 and 1: 10.500.
We can conclude that the survival rate for the Saloninus Augustus antoniniani fits in with the average values found by F. de Callataÿ. The survival rate of the Saloninus Augustus emission(s) does not indicate a successful demonetization. 

4. Coins for Saloninus Augustus in hoards 

Hoards inform us about the circulation and hoarding of the antoniniani for Saloninus Augustus. Hoards with coins for Saloninus Augustus can also be used to test the demonetization hypothesis.
As mentioned before, twelve hoards collectively contained a total of 19 antoniniani for Saloninus Augustus. Three of these hoards included two or more of these coins: Courcité [37] (two coins),
Stevenage [38] (three coins) and Eauze [39] (six coins). Some hoards closed shortly after Saloninus’ death, for instance Eauze (after 261 AD). Antoniniani for Saloninus Augustus can also be found in hoards closing much later, as Emneth [40] (after 271 AD) and Bonneuil-sur-Marne [41] (after 269-270 AD) will show.

A very late hoard is Coesmes [42], which has been dated in or shortly after 274 AD, but this terminus post quem is clearly too early. In this hoard, 80% of the coins for Tetricus I and Tetricus II are copies. In hoards buried around 274 AD, the share of copies for the Tetrici hardly even exceeds 7% [43].
The Coesmes hoard must be dated later: a relatively high share of radiate copies correspondents with a relatively late moment of burial [44]. Furthermore, looking at the share of Tetrici coins in the hoard (official and copies), we find that more than 75% of the analysed coins in this hoard was minted for these two rulers. Hoards buried before the death of Probus (282 AD) as a rule contain at least 45% of coins minted for the Tetrici, while the hoards buried later hardly even contain more than 20% of these coins [45]. From this data we can conclude that the Coesmes hoard most likely closed between 278 and 282 AD [46].
From Creil we have two hoards [47] from one owner. Creil I is a savings hoard, just like the hoards of Eauze and
Creil I includes coins from Julia Domna up to Postumus and one antoninianus for Saloninus Augustus. The hoard closed after 264 AD. Creil II contains coins with a low intrinsic value taken from circulation, ranging from Valerian to Tetricus II, with a 58% share for the coins of the Gallic usurpers. Creil II closes after 273 AD. ‘Good’ money was kept apart from ‘bad’ money by the owner of the two hoards. The coin for Saloninus Augustus (with a better silver content than the average coin in Creil II) was hoarded with the ‘good’ money.

If the owner had known the coin to be withdrawn from circulation, it would have ended up with the ‘bad’ money, not to be saved as a nest egg in hoard Creil I but kept for the time being to be spent on the next purchase together with the mass of other ‘bad’ coins from Creil II.
Other hoards with one Saloninus Augustus coin are from Cravent,
Evreux, Fins-d’Annecy, Mons-Boubert and Pommereuil, all in present day France.

To conclude, we have antoniniani for Saloninus Augustus in savings hoards, in late hoards and even in late savings hoards. We already saw that the survival rate of coins for Saloninus Augustus does not indicate a successful demonetization. Now we can also conclude that the presence of these coins in hoards pleads against demonetization.

4.1. The Eauze hoard

Especially the Eauze hoard (with six coins for Saloninus Augustus) should be noted. The coins for Saloninus Augustus in this much discussed hoard can also shed new light on the identity of the owner. The hoard consists of 28.003 coins, and a large quantity of gold jewellery. The jewellery, originating from the Rhine area and the share of coins from mints in that area led the authors of the publication on this hoards [48] to the conclusion that the owner must have had privileged relations, at least at a certain point in his career, with the north eastern part of Gallia and with Germania Inferior [49].
This can now be specified: the owner would have been in or near
Cologne in or shortly after 260 AD or had contacts there, in view of the six Saloninus Augustus coins from that city.

The occupation of the owner of the hoard has given rise to speculation in the past. Was he a notable citizen, burying his coins and his wife’s jewellery? Or was he a procurer or trader in female slaves as has been suggested? Judging from the coins, the owner could have been a (retired) army officer or a wealthy trader supplying the army. Many coins came straight from the mint and were uncirculated. The hoard included 2.586 coins with the Oriens Augg reverse for Valerian. Most of these coins (2.425) came from one obverse die (die duplicates). The owner must have been in close contact with the only source of new coins: the army. The six coins for Saloninus Augustus would imply the army on the Middle Rhine limes. 

4.2. Geographical distribution of the hoards 

Map 1 below shows the geographical distribution of all known hoards with a coin for Saloninus Augustus. Ten hoards were discovered in present day France, two in the United Kingdom. It is significant that no hoards with a Saloninus Augustus coin were found near Cologne, on the Middle Rhine or on the Moselle, areas otherwise very rich in coin hoards after 260 AD [50].


Map1. Hoards with one or more coins for Saloninus Augustus

Kaartje schatten met Sal Aug

1. Emneth
Stevenage (Hertfordshire)
3. Mons-Boubert
4. Pommereuil
5. Evreux
6. Cravent

7. Bonneuil-sur-Marne
8. Courcité
9. Coesmes
10. Creil
11. Eauze
12. Fins-d’Annecy

Many hoards with Saloninus Augustus coins were found in the north western part of France. We propose the hypothesis that the majority of the Saloninus Augustus coins were still in the possession of the soldiers which had defended Cologne against Postumus’ forces. After the city fell to Postumus, it would not have been wise policy to maintain the units that had been loyal to the imperial house in or near Cologne. A transfer of these troops (with Saloninus Augustus coins still in their purses) might be expected and the distribution of the hoards suggests a transfer to the heart of Gallia Lugdunensis, possibly near or even in Lutetia.
Of course, this hypothesis does not explain the outlying hoards: Eauze, Fins-d’Annecy,
Stevenage and Emneth.

5. Conclusion  

Saloninus, the emperor Gallienus’ son and Caesar in residence in Cologne, was captured and put to death after a siege by troops of the rebellious governor Postumus in 260 AD. The ancient written sources deal with the episode without mentioning the fact that Saloninus took the title of Augustus during the siege. However, a coin issue in two metals bears witness to this new rank. The Augustus title was not conferred by Gallienus to Saloninus, but adopted by him as an unplanned localized reaction to the rebellion and siege. The title was not confirmed by written sources or coins from other mints. 

In view of the small number of surviving coins, this issue was thought to be demonetization and melted down by Postumus after the siege. However, demonetization is a modern concept which should not be projected into antiquity without further analyses. Demonetization of ancient coins was a difficult and complex process, as has been demonstrated before.

After we determined the number of obverse dies of the Saloninus Augustus issue, based on a close die study, we calculated the original size and the survival rate of the issue, which closely conforms to the average survival rates established in a diachronic study. The survival rate does not indicate demonetization.  

When the incidence of Saloninus Augustus coins in hoards is reviewed, we also come to the conclusion that the demonetization hypothesis is less probable and should be rejected.


 (1) R. GÖBL, Die Münzprägung der Kaiser Valerianus I / Gallienus / Saloninus (253-268), Regalianus (260) und Macrianus / Quietus (260 / 262), MIR 36, 43, 44, Wien, 2000, p. 61.
(2) J. LAFAURIE, La Chronologie des empereurs gaulois, in RN 6.6, 1964, p. 91-127.

The Gallic Empire, Separatism and Continuity in the North-Western Provinces of the Roman Empire AD 260-274 (Historia Einzelschriften, 52), Stuttgart, 1987, p. 167.
(4) R. GÖBL,
op. cit. [n. 1], p. 99.
(5) N.SHIEL,
The Coinage of Saloninus as Augustus, in ANSMN, 24, 1979, p. 117-122.

Ein weiterer Antoninian des Saloninus Augustus, in Numismatisches NachrichtenBlatt 31, 1/1982, p. 6-7; N.M.McQ. HOLMES, Two new Antoniniani of Saloninus Augustus, in Seaby Coin and Medal Bulletin, dec. 1982, p. 363-364; H.H. GILLJAM, Acht weitere Antoniniane des Saloninus Augustus, in Geldgeschichtliche Nachrichten 21, no. 111, 1/1986, p. 10-11; H.H. GILLJAM, Weitere Antoniniane des Saloninus Augustus, in Geldgeschichtliche Nachrichten 21, no. 115, 9/1986, p. 248; H.H. GILLJAM, Antoniniane des Saloninus Augustus, in Mitteilungen der Österreichischen Numismatischen Gesellschaft, XXVII/6, 1987, p. 77-83; H.H. GILLJAM,
Saloninus Augustus- Neue Antoniniane seit 1987, in Mitteilungen der Österreichischen Numismatischen Gesellschaft, XXIX/6, 1989, p. 91-92; H.H. GILLJAM, Weitere Antoniniane des Saloninus Augustus, in Numismatisches NachrichtenBlatt 46, 12/1997, p. 617;
An Inventory of Romano- British Coin Hoards, London, 2000, p. 167-168, no. 723;
Drei weitere Antoniniane des Saloninus Augustus, in Geldgeschichtliche Nachrichten 43, no. 238, 5/2008, p. 153; S. SONDERMANN, Vier weitere Antoniniane des Saloninus Augustus, in Numismatisches NachrichtenBlatt 58, 4/2009, p. 158.
(8) N.SHIEL,
op. cit.[n.5], p. 120; H.H.GILLJAM, op.cit. [n. 7, 1982], p.6.
(9) N.M.McQ. HOLMES,
op. cit. [n. 7], p. 364.
(10) N. SHIEL,
op. cit. [n. 5], p. 120.
(11) R. GÖBL,
Antike Numismatik, München, 1978, p. 220, 238-239.
Ibidem, p. 52.
Ibidem, p. 52, p. 238-239.
(14) D. SMITH,
Die links, a Tool for the numismatists, in The Celator, December 1995, p. 12-17.

(15) Fig.1: GS 44, SONDERMANN 2009, ex CNG mail bid sale 78; fig. 2: GS 45, SONDERMANN 2009,
ex Klassische Münzen; fig. 3: GS 42, GILLJAM 2008, p. 153,
  no. 42; fig. 5: GS 8, GILLJAM 1987, p. 79-80,
coin 8.

(16) fig. 4: GS 43, SONDERMANN 2009, ex CNG mail bid sale 58.

(17) N. SHIEL,
op. cit. [n. 5], p. 119, coin no. 8 (GF 7).
(18) , retrieved April 30 2013.
(19) N. SHIEL,
op. cit. [n. 5], p. 120; H.H. GILLJAM, op. cit. [n.7, 1982], p. 6; N.M.McQ. HOLMES, op. cit. [n.7], p. 363; H.H. GILLJAM, op. cit. [n.7, 1987], p. 78.
(20) R. GÖBL,
op. cit. [n. 1], table 26, 917 f; RIC V, part I, 14.
(21) R. GÖBL,
op. cit. [n. 1], table 26, 916 c; RIC V, part I, 1.
(22) R. GÖBL,
op. cit. [n. 1], table 26, 916 f; RIC -
(23) R. GÖBL,
op. cit. [n. 1], table 26, 915 e; RIC V, part I, 13.
(24) Compare R. GÖBL,
op. cit. [n. 1], p. 170.
Some experiments in Greek minting technique, in NC 1963, p. 217-231.
(26) R. GÖBL,
op. cit. [n. 11], p. 237.
(27) F. de CALLATAŸ,
Les taux de survie des emissions monétaires antiques, médiévales et modernes. Essay de mise en perspective et consequences quant à la productivité des coins dans l’Antiquité, in RN 2000 6/155, p. 87-109; F. de CALLATAŸ, Quantifying monetary production in Greco-Roman times: a general frame, in Pragmateiai 2011, 19, p. 7-29.
(28) F. de CALLATAŸ,
op. cit [n. 27, 2011], p. 11, note 14.
(29) E.J.P. RAVEN,
The Amphictionic Coinage of Delphi, 336-334 BC , in NC 1950, p. 1-22.
(30) R. GÖBL,
op .cit. [n. 1], p.100.
(31) D. SCHAAD (ed),
Le trésor d’Eauze, éditions APAMP, Toulouse 1992.

(32) Zosimos, Nea Historia, 1.61.3: ‘He [Aurelian] likewise called in all the counterfeit money and issued new, to avoid confusion in trade’.

(33) A. KROPFF,
Late Roman coin hoards in the west: trash or treasure?, in RBN 2007, p. 73-86.
(34) A.A. BARRETT,
The Invalidation of Currency in the Roman Empire: the Claudian Demonetization of Caligula’s Aes, in G.M. PAUL (ed),
Roman coins and  public life under the Empire, Michigan, 1999, p. 83-93.
(35) F. de CALLATAŸ,
op. cit. [n. 27, 2000].
Ibidem, p. 99.
(37) G. AUBIN,
Le trésor de Courcité (Mayenne): Antoniniani et imitations de Gordien III à Victorin ,in
XI, 1989, p. 58
(38) R. BLAND,
Stevenage, Hertfordshire, in R. BLAND and A. BURNETT (eds), The Normanby hoard and other Roman coin hoards, London, 1988, p. 43-47.
(39) D. SCHAAD,
op.cit. [n. 31].

(40) A.S. ROBERTSON, An Inventory of Romano-British Coin Hoards, London, 2000, p. 167-168.

(41) J.- B. GIARD, Malicorne et Bonneuil-sur-Marne: deux trésors monétaires du temps de Victorin, in
6/8, 1966, p. 144-180.
(42) P. Le GENTILHOMME, Le trésor de Coesmes (Ille-et-Vilaine),in Gallia 1947, 5/2, p. 319-349.

Le trésor de Saint-Mard , Wetteren, 1970, P. 54-55.
Ibidem, p. 55-57; R. ZIEGLER, Der Schatzfund von Brauweiler, Köln and Bonn, 1983, p. 74.
(45) J.-P. CALLU,
La politique monétaire des Empereurs Romains de 238 á 311, Paris, 1969, p. 349-352;
op.cit. [n.43], p.55.
(46) A.C. KROPFF and J.P.A. Van der VIN,
Coins and continuity in the Dutch River area at the end of the third century AD, in European Journal of Archaeology 6(1), 2003, p. 69.
Le trésor monétaire de l’Ecluse de Creil, in Revue archéologique de Picardie, 1/1-2, 1985, p. 65-111.
(48) D. SCHAAD,
op. cit. [n. 31].
(49) Compare the review by P. GROS in
Gnomon 67/H.2, 1995, p. 178-180.
(50) Compare A. Kropff,
op. cit. [n. 33], p. 73, map 1.