FEL TEMP REPARATIO Falling Horseman Coinage Article

KOINON
The International Journal of Classical

Numismatic Studies
Volume II
Nicholas J. Molinari, General Editor
Shawn Caza, Associate Editor
Lloyd W.H. Taylor, Associate Editor
Produced by the
Societas De Tauro Cum Facie Humana
Archaeopress Publishing
Oxford 2019
Editorial Advisory Board
All the articles appearing in this journal have gone through a thorough process of blind peer-review. I am greatly indebted to the
members of the editorial advisory board, all of whom graciously volunteered their expertise for the preparation of this journal.
Alberto Campana, IT Mark Fox, US David Sear, US

Shawn Caza, CA József Géza Kiss, HU Andrew Short, CA
David M. Chico, ES Bob Langnas, US Nicola Sisci, IT
Victor Clark, US Vincenzo La Notte, IT Lloyd W. H. Taylor, AU
Curtis Clay, US David MacDonald, US Joseph Uphoff, US
Phil Davis, US Gavin Richardson, US
Tjaart de Beer, CH Martin Rowe, SE
Communications for the editors, manuscripts, and books for review should be addressed to the
General Editor. All enquiries to:
Nicholas J. Molinari, General Editor

1784 Providence Road,
Northbridge, Massachusetts
njmolinari@gmail.comw
https://koinonjournal.wordpress.com/
ISSN 2631-5874
ISSN 2631-5882 (e-pdf)
ISBN 978-1-78969-355-3
ISBN 978-1-78969-356-0 (e-pdf)
Published by Archaeopress Publishing Ltd, Oxford, UK
Subscriptions to KOINON should be sent to

Archaeopress Publishing Ltd, Summertown Pavilion, 18-24 Middle Way,
Oxford OX2 7LG, UK
Tel +44-(0)1865-311914 Fax +44(0)1865-512231
e-mail info@archaeopress.com
http://www.archaeopress.com
Σωκράτης: ὦ φίλε Πάν τε καὶ ἄλλοι ὅσοι τῇδε θεοί, δοίητέ μοι καλῷ γενέσθαι τἄνδοθεν: ἔξωθεν δὲ
ὅσα ἔχω, τοῖς ἐντὸςεἶναί μοι φίλια. πλούσιον δὲ νομίζοιμι τὸν σοφόν: τὸ δὲ χρυσοῦ πλῆθος εἴη μοι ὅσον μήτε
φέρειν μήτε ἄγειν δύναιτο ἄλλος ἢ ὁ σώφρων.
ἔτ᾽ ἄλλου του δεόμεθα, ὦ Φαῖδρε; ἐμοὶ μὲν γὰρ μετρίως ηὖκται.
Φαῖδρος: καὶ ἐμοὶ ταῦτα συνεύχου: κοινὰ γὰρ τὰ τῶν φίλων.
-Plato, Phaedrus (closing prayer), 279b-c

Contents
An Introductory Note from the General Editor, with Recourse to Plato and Eukleidas ��i
GREEK NUMISMATICS
Numismatic evidence (or not) for the aphippodroma horse race at Larisa��1
Rosanagh Mack
A Bacchid at Apollonia: a late survival of an ancient family��20

David Macdonald
An unusual depiction of Ba‘al Arwad and a hippocampus on coins of Arados during the
Persian Period��23
Martin Rowe
The Macedonian Mint At Susa (319/8-312/1 BC)��28

Lloyd W. H. Taylor
The Susa wreath group Alexanders: The first step in the transformation of an anchor seal
to a dynastic emblem��63
Lloyd W. H. Taylor
A discussion on provenance research with some early provenances uncovered��83

John Voukelatos
ROMAN NUMISMATICS
The Youthful God revisited: Veiovis on Roman Republican denarii��97

Tyler Holman
An enigmatic denarius of M. Herennius��109

Phillip Davis
Some further ideas on a double-obverse bronze of the Constantinian period from the
Antioch excavations��111
Shawn Caza
Back in the saddle again: a re-examination of the FEL TEMP REPARATIO Falling
horseman type��113
Shawn Caza
v
BYZANTINE AND RELATED COINAGES
The ‘Sirmium Group’ – an overview��147

Dirk Faltin
MEDIEVAL AND EARLY MODERN COINAGE
Numismatic letterforms of ‘A’ in medieval Europe: A classification system��176

David B. Spenciner and Marina V. Spenciner
Did Louis X of France mint deniers tournois? (Notes on a few deniers tournois à la croisette)��188
Andrei Bontas
A CATALOG OF NEW VARIETIES��198
vi
Back in the saddle again:
a re-examination of the FEL TEMP REPARATIO Falling
horseman type
SHAWN CAZA
INTRODUCTION
The FEL TEMP REPARATIO Falling horseman bronze coin type (FTR-FH) was introduced in 348 AD by
Constantine I’s two surviving sons, Constans and Constantius II, as part of a three-denomination
series of coins, all with the same reverse legend.1 Over the ten years that it was minted, the FTR-
FH type was struck in large numbers for Constantius, and his two Caesars, Constantius Gallus and
Julian, and in much smaller numbers for Constans and the usurper Magnentius.
The basic design of the FTR-FH type is simple. On the right side of the reverse, a large
armoured figure stands thrusting a spear downward. On the left side, a small horse tumbles forward
while its rider is in the process of falling off, or, on some examples, has just fallen off (hence the use,
by different authors, of ‘Falling Horseman’ and ‘Fallen Horseman’).
Despite this relatively simple design, this is one of the most complex coin types of the late
Roman period. The type was struck in many different sizes and weights. It bears a bewildering
variety of marks in the obverse and reverse fields. In addition, there is significant variation in the
design -- the position of the horseman, horse, and the Emperor all vary, as does the headgear of the
horseman. Furthermore, there are differences of opinion regarding the dating of the type and the
meaning of its imagery.
In an attempt to make sense of this often misunderstood reverse type, this article will propose
a division of the type into six sequential series, and into sub-types and variations based on the
position and design of the horseman. It will also offer a new schema for the sub-types and designs.
Though other taxonomic schemes have been attempted by previous authors, I believe that this new
outline will better assist with understanding the chronology of this type and therefore assist with
re-constructing the metrology and economics of the time.2
1
This legend is abbreviated. Neither FEL nor TEMP represent proper Latin words in this context. There have been
various theories about what the full legend is, and therefore what it translates to. Stevenson (1889) proposed FELIX
TEMPORVM REPARATIO. Mattingly (1933) agreed with this, which he translated as ‘A happy renewal of times’, though
he also posited FELICIVM TEMPORVM REPARATIO as an option. Jones (1990) followed this latter, which he translated as
‘Restoration of happy times’. Failmezger (2002) used the same translation but also provided a more colloquial translation
‘Happy days are here again’. Ferrando (2010) proposed FELICITATIS TEMPORVM REPARATIO, which he translated as ‘Le
retour des temp heroux’, which can be rendered in English as ‘The return of happy times’. Most recently, Abdy, in Metcalf
(2012), went back to FELIX TEMPORVM REPARATIO with the translation ‘Return of fortunate times’. Determination of the
exact expansion, and therefore proper translation, is likely unresolvable.
2
This article is focussed on the reverse type. It does not provide full details on which issues were struck for which
individuals, on obverse details such as the different diadems, or on the obverse or reverse legend breaks. Other works,
such as RIC-VIII, should be consulted for such details.
KOINON II 2019: 113–146

SHAWN CAZA
NUMISMATIC HISTORIOGRAPHY
In his 1933 Numismatic Chronicle article ‘FEL. TEMP. REPARATIO.’, Harold Mattingly noted that ‘we
have not even a preliminary study, defining [the] general scope and intention’ of this coinage.3
Mattingly was correct - no earlier scholar provides any insight into the structure or dating of the
FTR series. Mattingly himself was primarily concerned with the meaning and purpose of the FTR
series and its introduction. His schema proposed only two sizes as part of the initial FTR series,
using the terms M.B. and P.B. that were common at the time.4 He noted that an intermediate size
was found, bearing the mark LXXII at Aquileia, ‘perhaps denoting the number of pieces struck to
the pound’. He also concluded that the smaller FTR-FH were probably later than the larger, and
that FTR-FH was the only FTR type that commonly continued in circulation.5
The next important contribution to the study of the FTR-FH came in Carson, Hill and Kent’s
Late Roman Bronze Coinage (hereafter LRBC) of 1960, based on Numismatic Chronicle articles of 1957.
Though LRBC incorrectly dated the start of the FTR series to 346, it did identify correctly that there
were three different sizes in the initial FTR set instead of Mattingly’s two. LRBC referred to these as
the AE2, reduced AE2, and AE3. LRBC divided the FTR-FH type into five variations, although there
were really only four variations given that their variation 2A differs from variation 2 only in having
a star in the field.6
In The Roman Imperial Coinage Volume VIII (RIC-VIII), Kent accepted Mattingly’s 348 start date
and reduced the number of FTR-FH variations to three. He also identified varieties for the horseman
design based on the hair and clothing, but did not apply this to all mints.
Bastien made important contributions to the study of the FTR-FH type in his 1985 work on the
coinage of Lyon. He was, however, unable to fully sort out the confusing array of field marks, and
missed the weight decrease of the final ‘M’ issue. Both Carson (1990) and Depeyrot (1992), on the
other hand, did note this further reduction. A summary of the scholarship on the FTR-FH type, up to
1987, was included in R.J. Brickstock’s important study of the FTR copies from Britain.
Failmezger divided the FTR-FH type into consecutive series based on the symbols in the
fields, and on the size reductions. He presented five series in his 1992 article in The Celator, and six
series in his 2002 book.7 While I do not agree fully with Failmezger’s division or dating, his work
was an important contribution to the understanding of the chronology of the type.
Failmezger’s 1992 Series Failmezger’s 2002 Series

1st 348 to 350 blank or A 1st 348 to 350/1 blank or A
2nd 350 to 352 Γ 2nd 350 to 352 B
3rd 352 to 354 various + Γ, (AE3) 3rd 350 to 352 III or Γ
4th 354 to 357 many 4th 353 Δ, S or LXXII
5th 357 to 360 AE4 size 5th 354 E and others
6th 355 to 361 M
3
Mattingly 1933 (1977 reprint) p.3.
4
Brickstock 1987, p.11, ‘His labels are derived from an early catalogue of types by H. Cohen [Cohen, 1880-1892], and stand
for ‘moyen bronze’ [middle bronze] and ‘petit bronze’ [small bronze], respectively.’
5
Mattingly 1933 (1977 reprint) p.9-10 and 15.
6
LRBC (1978 reprint), p.41 and 108.
7
Failmezger (1992), 18 and (2002), 112 - 119.
114
Back in the saddle again
THE BRONZE COINAGE REFORM OF 348 - THE FEL TEMP REPARATIO DENOMINATION SET
A few words need to be said about the broader FEL TEMP REPARATIO set before examining the
history of the FTR-FH type. In 348, Constans and Constantius II replaced the centenionalis, which
had been the only regular bronze denomination since 318, with a new three-denomination set, all
of which carried the reverse legend FEL TEMP REPARATIO.8
This new denominational set was innovative - each of the three denominations was
distinguished by size (weight and diameter) and by the obverse bust style. Each denomination
was also issued with two distinct reverse types. Each reverse type was struck for both Emperors -
Constans and Constantius II. However, they were not struck in equal quantities for the two Emperors.
For example, in the large denomination, the FTR-FH was issued primarily for Constantius, while the
Emperor on galley type was issued primarily for Constans.
Although the FTR types were clearly intended as a set of related denominations, it seems that
they did not all appear simultaneously. There is some debate about the exact order in which the
denominations and types were introduced. The most common scenario is that they were introduced
from smallest to largest, but the mint mark evidence is not entirely clear. Furthermore, it appears
that Constans’ types appeared before Constantius’, thus the Galley type appeared before the FTR-FH
type. This dating difference indicates that the FTR series originated in Constans’ Western half of
the Empire.
Denomination Weight Standard Bust Faces Constans’ Type Constantius’ Type
Large 1/60 L Right Large Galley Falling Horseman

(maiorina)
Medium 1/72 L Left Hut Emperor with Two Captives
(centenionalis)
Small 1/120 L Right Small Galley Phoenix
(1/2 centenionalis)
However, the FTR three-denomination set did not remain in use for long.9 The middle denomination
ceased to be struck by the second half of 349 in the West and by early 350 in the East. The small
denomination ceased to be struck by mid-350 at most mints, though it lasted into early 351 at a few.
As the large Galley type also ceased to be struck in early 351, the FTR-FH was left as the sole FTR type
in Constantius’ territory.
8
This centenionalis coinage consisted of a single bronze denomination, struck at 1/192 of a Roman pound (hereafter L, for
librum), theoretically 1.69 g. Several reverse types were struck in this denomination between 341 and 348: most
importantly the VICTORIAE DD AVGG Q NN Two Victories standing holding wreaths type struck in the West, and the Vota
types (VOT XX MVLT XXX and its Antioch variation VOT XV MVLT XX) struck in the East.
9
It is unlikely that maiorina, cententionalis and half-cententionalis were the original names for these denominations.
Maiorina simply means large – as in ‘the largest coin’. It must also be noted that in addition to these six main types, there
were a few other rarer FEL TEMP REPARATIO types. A medium denomination Emperor riding down two barbarians type
was struck for Constans and Constantius at Rome from mid-349 to February 350. A large denomination Emperor holding
labarum type was struck for Magnentius at Lyons in early 350. A large denomination Emperor standing left holding victoriola
and labarum, stepping on captive type was struck for Constantius and Gallus at Thessalonica from May to July 351. Finally, a
very rare small denomination Large M with three dots type was struck for Constantius at Rome in 356.
115
SHAWN CAZA
The remaining FTR-FH type continued to be struck, but shrank in weight and size several
times. It went through a total of six series before being replaced by the SPES REIPVBLICE type in
358.10 These six series are detailed in the next section. The accompanying charts incorporate coin
issues listed in RIC-VIII, LRBC and mint-specific works, such as Bastien (1985), Depeyrot (1996, 2001),
Ferrando (2010) and Sagramora (2010). However, not every issue listed in these sources has been
included. Issues noted in RIC-VIII as requiring confirmation have been included only if more than
one example has been confirmed. Similarly, issues from any of these sources that are based on a
unique example, or two, have not been included unless other examples have been found. While a
unique coin could indicate an otherwise unknown issue, given the ubiquity of this type, it is more
likely to represent an error or a misreading due to wear or corrosion.
THE SIX FEL TEMP REPARATIO FALLING HORSEMAN SERIES
The first series: April 348 to mid-349
It is now widely accepted that the FEL TEMP REPARATIO coinage was introduced in 348 as part
of the celebration of Rome’s 1,100th anniversary on 21 April 348. This date was first proposed by
Seeck in 1890, though the best known explication of the link to the 1,100th anniversary celebrations
is found in Mattingly’s 1933 work on the FEL TEMP REPARATIO series.
The date of 348 was not, however, universally accepted. In the early part of the 20th century,
Voetter proposed ‘344 or 345’ for the start of the FTR coinage.11 In LRBC, Carson, Hill and Kent
proposed 346.12 However, by the publication of RIC-VIII, Kent had adopted the 348 dating, as did
Carson in his 1990 book.13 Nevertheless, the continued utility of LRBC as a reference means that the
start date of 346 continues to appear in some catalogues and websites.
Attempts to place the start of the FTR coinage much earlier than 348 run into problems
related to the coinage of the previous period. The small VICTORIAE DD AVGG Q NN coinage,
introduced in 341, is clearly not part of the FTR series and, given it was only struck for Constans
and Constantius, must pre-date it. However, to limit that coinage to only three, four or five years
condenses a significant volume of coinage, and a large number of mint and field marks, into a
short period.14 Such a dating would also force the contemporary type struck for Constans, VOT XV
MVLT XX, to be struck well before, but not during, his actual Vota celebrations, which were likely
celebrated sometime between 346 and 348.15 At many mints, especially in the West under Constans,
the FTR-FH type was introduced after the FTR Galley type, even though the Galley type was struck
for both Constans and Constantius.
10
For the purposes of this article, a series is a division of the coinage based on changes in reverse type, weight standard
and/or significant markings. A series is a subjective grouping made by modern authors. An issue is a group of coins
that all bear the same mint and field marks. An issue might, nevertheless, include coins struck for several different
personalities and/or might include several different reverse types. Unlike a series, an issue is clearly an organizational
grouping intended by Roman mint officials.
11
Voetter (1921).
12
Carson, Hill and Kent provide no explanation of why they date the appearance of the FTR coinage to 346.
13
Kent (1981), 34 and Carson (1990), 179.
14
Kent (1981), 51. According to RIC-VIII, the VICTORIAE DD AVGG Q NN Two victories type was struck in two issues at
Thessalonica, seven at Siscia, eight at Arelate and Aquileia, nine at Trier and Lyons, and as many as sixteen at Rome.
15
Kent (1981), 34.
116
The FTR-FH coins of this first series lacked obverse field marks, although reverse field marks
are common. The symbol ✶ appears in either the exergue or the reverse field, on many issues in
this series, although most eastern mints began with an issue with no reverse field mark.
The FTR-FH coins of the first series were struck at a theoretical weight standard of 1/60
L (5.40 g).16 Average weights were, as is usual with almost all ancient coinage, lower than the
theoretical weights. As many sources do not distinguish the FTR-FH type by series, it is difficult
to find statistics that treat the first and second series individually. It appears that there was little
decline in the average weight between the first and second series, but this is uncertain. The average
weight for the two series, treated together, is 5.2 to 5.3 g. There was, however, a wider variation
in the weight of the early FTR-FH coins than was the case for the coinage of the preceding few
decades. For example, in the Boljetin Hoard the weights of the two largest FTR denominations
blur into each other at many mints.17 This wide weight distribution is likely a sign that the mints
encountered technical challenges minting the largest denomination.
One problem might have been their size, which was larger than any bronze coin that had
been minted for 39 years (in other words, for much longer than a generation of mint workers).18 In
addition, the alloy being used for the new FTR-FH coins had a high lead content, which was not the
case with the heavy coinage of the late third and early fourth centuries. More importantly however,
as Cope noted, the weight standard itself might have proved challenging as it was difficult to cut
a libra strip of bronze into 60 even pieces during the manufacturing process. Unlike some other
weight standards, this required cuts into five pieces, as well as into two or three, and was therefore
more difficult to accurately estimate.19 It is possible that it was actually made at 1/64 L (5.06 g) at
some mints. This was an extremely simple fraction involving only halving-cuts and would explain
the low weights. However, use of a 1/64 L weight standard would also mean that a great many
surviving coins are overweight. The pearl ring diameter (PRD) of the large denomination FTR-FH
coins of the first series was 21-22 mm.20
At its introduction in 348, the fineness level of the FTR-FH type appears to have been around
2.8% (which likely represents 8 scrupulae of silver added per pound of bronze). However, figures for
this denomination pose problems as the fineness level appears to be different in the Western and
Eastern parts of the Empire. The figure of 2.8% appears to apply only to the first series as it was
16
Unless otherwise noted, weight and fineness standards throughout this article are derived by examining several
sources including Bastien (1985), Carson (1990), Cope (1974), Depeyrot (1992 and 2001), Kent (1981), and Vasic (1978). Of
these, Cope, Depeyrot and Bastien are more detailed and useful than Kent’s RIC-VIII, if less well known. Full details of
how these figures were arrived at would require another, equally lengthy article. It is important to note that while the
weight standards are based on large samples, and are likely fairly accurate, the fineness standards are based on far fewer
samples, all of which suffer from the biases inherent in the various mensural technologies used. Therefore, less certainty
should be ascribed to the fineness figures and to any conclusions derived from them, such as the nominal value of the
coins. Nevertheless, due to the nature of the studies, the overall relationships, patterns, and the degree of change over
time, are all likely to be fairly representative of the actual situation, even if the representativeness of specific figures are
less certain.
17
Vasic (1978), 140.
18
Since the fall of the nummus from 1/48 L to 1/72 L in late 309.
19
Cope (1974), 231.
20
The only official, or planned, diameter for a coin was the diameter of the die used to strike the coins. The actual
diameter of the coin body, or flan, was an ‘accident’ of production. This is because the final coin size will depend on
the size of the blank flan and on the striking force used, whereas the diameter of the die is easy to control and remains
standard for each issue. The official diameter is therefore best found by measuring the diameter of the ring of dots near
the outer edge of the coin. This diameter is known as the pearl ring diameter or PRD.
117
SHAWN CAZA
struck in the West. The fineness level appears to have dropped to 2.45% or 2.1% (7 or 6 scrupulae)
for the second series in the West, and was as low as 1.4% (4 scrupulae) for the first series in the East.
In other words, it appears that Constantius II did not see fit to strike the large denomination at the
higher fineness level introduced by his brother Constans.
The FTR-FH type appears to have been silvered when it was introduced, though the silvering
was thinner than that used on coins of the earlier fourth century, and rarely survives on FTR-FH
coins today. This silvering appears to have lasted through the first four series of FTR-FH coins,
though perhaps only sporadically on the fourth, and then to have ceased by the fifth series.
The special globe bust
A special, rare obverse bust was used for a single series of FTR-FH coins at three different mints.
Commonly known as the ‘globe bust’, this portrayed the Emperor in pearl diadem, draped and
cuirassed bust facing right and holding a globe. Though similar to the usual bust for the medium
FTR denomination, where the Emperor also holds a globe, this special globe bust is facing right, like
the regular large denomination bust, instead of left like the medium denomination bust.
The special globe bust first appears on the FTR-FH type struck during the second issue of
the first series at Thessalonica (with TSA❋ mint mark), and is found for both Constantius and
Constans. This represents the initial issue of the FTR-FH type at Thessalonica, as the type was not
struck at that mint during the first issue of the first series.21 In addition to Thessalonica, the special
globe bust was also struck at Rome, where it appears on the ❋|- RP issue for both Constantius II
and Constans. These issues, at Rome and Thessalonica, were likely struck towards the end of the
first series, in other words, during the first half of 349. This special globe bust was also used on the
rare series of FTR-FH coins struck at Rome for Constantius under Magnentius, dating from January
to February 350.
While such a clear and unique bust variation must have signified something, its purpose
remains unknown today. Neither Constans nor Constantius were consuls at that time, nor did they
hold any other unique powers.22 Constantius was in the East throughout this period. Constans,
on the other hand, was in Sirmium in the Balkans in May 349, so it is possible that he visited
Thessalonica around the time that the special bust was struck there. However, even if Constans
chose to strike a special bust for both himself and his brother while visiting Thessalonica, it is
unclear why it was struck only there and at Rome, and then why it would be struck again at Rome
for Constantius by Magnentius shortly after Constans’ murder approximately half a year later.
21
The Galley type was the only large denomination struck during the first issue of the first series at Thessalonica. It was
struck with the TESA mint mark.
22
The two brothers were consuls together in 346, and Constantius served as consul several more times, from 352.
118
Example of FTR-FH from first series with special Globe Bust, Thessalonica (-|- TSΔ✶).
AMB TR LUG ARL RO AQ SIS SRM TH HER CON NIC CYZ ANT ALE
-|- -|- ✶ |- -|✶ -|✶ -|- -|- -|- -|- -|- ✶ |- -|-
PLG✶ PARL✶ RP AQP ASIS TSA✶ SMHA CONSA SMNA SMKA ANA ALEA
-|✶ -|✶ ✶ Γ|- ✶ |- ✶ |- ✶ |-
PLG AQP● |-SMHA CONSA SMNA SMKA ALEA
Γ|- -|✶
CONSA✶ SMNA
-|-
SMNA✶
Chart showing field marks on issues struck by each mint during the first series, April 348 to mid-349.23
Notes: While Kent indicated in RIC-VIII that the two issues for Aquileia listed in the above chart may belong to
the period of Magnentius, it is clear that they belong here in the first series. They lack the A on the obverse,
which was introduced before Magnentius’ usurpation at all Western mints.
The second and third issue of the FTR-FH at Constantinople, had a Γ|- reverse field mark and were struck for
Constantius only. A parallel issue of the FTR Galley type was struck at the same time for Constans with reverse
field mark S|- instead of Γ|-. That these Constantinopolitan field marks are not officina marks is made clear
by the fact that each of these field marks appears on coins with a range of different officina marks in the
exergue.
Kent lists issues with F|- at Nicomedia, but notes that this field mark required confirmation. While a few
examples of this coin are known, it is almost certainly an engraving error for a Γ from the third series and so
is not listed here.24
23
In this and following charts, the information in the cell begins with a description of any marks found in the obverse
field, followed by reverse field marks (the | divides the field into left and right, thus ✶|- means star in left field, right field
blank), and ends with the mint mark found in the exergue, at the bottom of the reverse. While the issues in the chart are
placed in sequential order, as far as can be deduced, the issues across each row should not be considered as simultaneous
issues at the different mints. The exact dating of each issue within each series is uncertain. Mint abbreviations are: AMB
(Ambianum/Amiens); TR (Treveri/Trier); LUG (Lugdunum/Lyons); ARL (Arelate/Arles); RO (Rome); AQ (Aquileia); SIS
(Siscia); SRM (Sirmium); TH (Thessalonica); HER (Heraclea); CON (Constantinople); NIC (Nicomedia); CYZ (Cyzicus); ANT
(Antioch); ALE (Alexandria).
24
Kent (1981), 476.
119
SHAWN CAZA
The second series: mid-349 to the end of 350
The second series of FTR-FH coins is distinguished by the presence of an A located behind the
bust in the obverse field. Coins from this series almost always had a symbol on the left side of the
reverse field. This was usually an A, although a chi-rho (☧), B and Δ were used more rarely. The
second series of FTR-FH was not struck at Constantius’ eastern mints (from Heraclea to Alexandria).
The second series maintained the same weight standard as the first, 1/60 L (5.40 g). The PRD also
remained the same at 21-22 mm.
A on A|✶ A on A on A on A on A on
Obv, PLG Obv, Obv, Obv, Obv, Obv,
A|- A|- A✶ |- ☧ |- A|- A|-
TRP PARL RP AQP ●ASIS● TSA
A on A on A on A on A on A on
Obv, Obv, Obv, Obv, Obv, Obv,
A|- A|- A✶ |- A|- A|✶ B|✶
PLG PAR R●F●P AQP● ●ASIS● ●TSA●
A on A on A on A on
Obv, Obv, Obv, Obv,
A|- A|● ✶|- Δ|✶
✶ PLG PAR AQP● ●TSA●
A on A on A on
Obv, Obv, Obv,
A|- -|- B|-
FPLG PAR✶ ✶ |-
branch
AQP
branch
A on
Obv,
A|L
PAR or
PARL
A on
Obv,
A|-
PAR●F
A on
Obv,
A|-
PAR
branch
Chart showing field marks on issues struck by each mint during the second series, mid-349 to the end of 350.
Notes: Bastien notes that the issue from Lyons with an A in the reverse but not on the obverse, which is very
rare, likely represents limited issues struck by these mints at the beginning of the second series in which
these mints misinterpreted the instruction to place the A on the obverse. The same may be true of the single
specimen from Arelate referenced by Ferrando with A|- PAR branch on the reverse but no A on the obverse.25
Arelate appears to have struck a great profusion of issues during this series. In addition to the issues listed in
the chart, several other markings are known, mostly stemming from unique examples: A|- PAR●; A|● PAR●L;
A|✶ PARL. It is uncertain if any of these are official issues. In any event, most of the Arelate coins were struck
in early 350 under Magnentius, though in the name of Constantius. While the mark PARL was introduced on this
coinage in 349, the mark PAR was introduced only after Magnentius’ takeover of Gaul in early 350.
For Aquileia, I have reversed the order of the first two issues compared to that used in RIC-VIII. This is
because, while the issue with the chi-rho is found for Constans and Constantius, the issue with the A on the
reverse is found for Constantius and, on another contemporary coin type, for Magnentius.
The second and third issues at Thessalonica were struck for Constantius under Vetranio.
25
Bastien (1985), 53 - 54; Ferrando (2010), 244.
120
The FTR-FH coins of the second series appear to have been struck at a slightly lower fineness level
compared to the first series: 2.1% or 2.45% instead of 2.8%. This second series is likely linked in
some way to the law of 12 February 349 prohibiting the melting down of bronze coins to remove
their silver. While the introduction of such a law is a sign that there was likely an Empire-wide
problem, lowering the fineness standard of the coinage would have been a complementary, and in
fact better, solution to the problem than the law itself was. The second series may therefore have
had a lowered fineness level in an official strategy of attempting to deter the further melting of
the coins. The new coins would have contained less silver, and therefore would have had a higher
fiduciary value content. If this were the case, the new markings, the A on the obverse, would have
helped mint officials to identify which coins to keep in circulation, and which to withdraw in order
to re-melt so as to remove the excess silver for the benefit of the state.
We can, therefore, probably date the start of the second series to shortly after the February
law, likely by mid-349. This dating is supported by the fact that this series was clearly struck for
some time at Trier, Lyons and Arelate (where there were six different issues) before these mints fell
to Magnentius in January 350.
Example of FTR-FH from the third series: Second issue

from Thessalonica (B|✶ ●TSB●).
The FTR-FH Type under Magnentius: 350 to 351
The usurper Magnentius struck the FTR-FH type for himself during his first coinage phase (18
January to 27 February 350), but only at Arelate.26 These coins are very rare today and must have
been struck in very limited numbers.
Mint marks show that Magnentius also struck the FTR-FH type for Constantius during a few
different phases at several mints. There is, however, some confusion regarding these issues for
Constantius. At times, the mint marks make it difficult to determine exactly when the coins for
Constantius were struck and, therefore, whether they were struck while Constans was still alive or
26
The dating of Magnentius’ coinage phases, as used in this article, is taken from my previous article, Caza (2018), which
incorporates the changes made to Bastien’s original schema that are necessary to accommodate the re-dating of
Nepotian’s usurpation from 350 to 351.
121
SHAWN CAZA
under Magnentius. Our two main sources, Kent and Bastien, do not fully agree in their analysis of
the evidence.
At Lyons, where Magnentius struck a new and unique FTR type for himself (see footnote 8),
neither Kent nor Bastien believed that Magnentius struck any FTR-FH coins for Constantius. I have
chosen to follow them despite the fact that the mint mark A|- FPLG was used under both Constans
and Magnentius, and that it is therefore at least theoretically possible that some of the FTR-FH
coins for Constantius with the A|- FPLG mint mark were struck under Magnentius.
At Trier, the FTR-FH type is rare for Constantius. Kent believed that all examples for
Constantius were actually struck by Magnentius at the same time as he struck the large Galley type
for himself.27 By contrast, Bastien believed that all of Constantius’ FTR-FH coins were struck by
Constans prior to Magnentius’ usurpation, and that none were struck by Magnentius. The mint
mark A|- TRP is ambiguous as it was used on the large Galley type by Constans before Magnentius’
usurpation, and by Magnentius during his first coinage phase. Coins for Constantius could,
therefore, have been struck under either, or both. This means that we will likely never know who
exactly struck coins for Constantius at Trier. However, given the situation at Lyons, as described
above, I have chosen to follow Bastien in assuming that Magnentius did not strike any of the FTR-FH
coins for Constantius and that they were in fact struck by Constans in the second series just prior
to Magnentius’ usurpation (i.e., during the second half of 349).
At Arelate, which, as noted above, is the one mint where Magnentius struck the FTR-FH
type for himself, Kent and Bastien again differ about whether Magnentius struck the FTR-FH for
Constantius. Kent did not believe that any FTR-FH coins were struck for Constantius by Magnentius
during his first phase, despite the fact that Constans’ mint mark A|- PARL continued to be used by
Magnentius and that, therefore, some of the FTR-FH coins for Constantius with this mark could
theoretically have been struck under Magnentius. Bastien, by contrast, believed that the mint
mark evidence showed that Magnentius did strike some of those coins for Constantius. Both
authors agreed, based on mint mark evidence, that Magnentius definitely struck the FTR-FH type
for Constantius during the third phase (January to August 351), at the same time that Magnentius
struck the GLORIA ROMANORVM Emperor mounted type for himself (the issues with mint mark PAR).
Unfortunately, as is the case for Trier, we will likely never know who is correct regarding the first
phase.
At Rome, both Kent and Bastien agree, based on the mint marks, that Magnentius struck the
FTR-FH type for Constantius during his second phase (27 February to the end of 350). At Aquileia
the situation is the same as for Rome with respect to the second phase, though there the FTR-FH
type continued to be struck for Constantius during Magnentius’ third phase, with the reverse mark
A|- AQP●.
27
Kent (1981), 132.
122
MINT STRUCK FOR STRUCK FOR CONSTANTIUS II BY MAGNENTIUS

MAGNENTIUS (Magnentian coinage phase)
Trier No Possibly, but not likely.
Lyons No Possibly, but very unlikely.
Arelate Yes Possibly in Phase 1. Yes during Phase 3 (January - August 351).
Rome No Yes, during Phase 2 (27 February - end 350).
Aquileia No Yes, during Phase 2 and 3 (27 February 350 - August 351).
The FTR-FH Type under Vetranio: 1 March to 25 December 350
This chaotic period saw a second usurpation shortly after, and in response to, Magnentius’
usurpation. On 1 March 350, Vetranio (Constans’ military commander in the Balkans) declared
himself Emperor in order to prevent Magnentius from taking control of the Balkans. Vetranio
remained loyal to Constans’ family and turned over all power to Constantius when the latter arrived
at Naissus (Nis, Serbia) on 25 December 350.
During his usurpation, Vetranio struck three issues of the FTR-FH type for Constantius
at Thessalonica. These were struck while Vetranio struck new types (VIRTVS EXERCITVM and
CONCORDIA MILITVM) for himself. Vetranio did not strike the FTR-FH for himself and hence there
are no genuine issues bearing his name.
The first FTR-FH issue under Vetranio was struck while he struck the VIRTVS EXERCITVM
type for himself and cannot be distinguished from FTR-FH coins struck before Vetranio’s usurpation
as they bear the same mint mark, A|- TSA. However, the next two FTR-FH issues, struck while
Vetranio struck the CONCORDIA MILITVM type for himself, can be distinguished by their unique
mint marks, B|❋ ●TSA● and Δ|❋ ●TSA●, which were not used on other FTR-FH issues.
The third series: beginning of 351 to September 352
The second series of FTR-FH coins was replaced in the West by Magnentius’ own new coin types
in early 350. However, in the eastern part of the Empire, outside Magnentius’ control, a new third
series began to be struck in early 351.
The start of this series is often dated to the beginning of Gallus’ reign as Caesar on 1 March
351. This is because the coinage is found for both Constantius and Gallus. There is no reason,
however, for the introduction of the new series to be linked to Gallus’ accession. It is more likely
that this third series was introduced at the start of 351, while Constantius was ruling alone. This
was shortly after Constantius took over control of the Balkans from Vetranio on 25 December 350,
and a January date would better match the Roman tradition which often saw the introduction of
coinage changes at the start of the year.
Coins of the third series bear a standard mark in the reverse field. The mark used was a III
at Constantius’ central mints of Rome, Siscia and Sirmium, and a Γ at Constantius’ eastern mints
(from Thessalonica to Alexandria). Both of these marks clearly signify a third series, being the
Roman and Greek numerals for three, respectively. Interestingly, Constantius’ three central mints
maintained the obverse A of the second series on this third series.
123
SHAWN CAZA
The third series was struck at the same 1/60 L weight standard as the first two series, though
with a slightly reduced weight average of circa 5.0 to 5.1 g.28 The PRD also remained the same as
for the first two series at 21-22 mm. The fineness level fell to approximately 1.05% (3 scrupulae)
with the introduction of this series. This may be the reason that all of the coins of this series were
clearly marked - to allow mint officials to identify them and withdraw the earlier series, which
contained more silver. However, if this were the case, it is uncertain why the mark was moved from
the obverse to the reverse.
The third series appears to be more plentiful than either of the first two series despite the
fact that it was not struck at the Western mints. It was especially common in the Eastern part of the
Empire, where its issuance coincided with preparations for the Persian wars.
A on A on A on Obv, Γ|- Γ|- Γ●|- Γ|- Γ|- Γ|- Γ|-
Obv, Obv, III|- ✶ TSA✶ SMHA CONSA SMNA SMKA ANA ALEA
III✶|- III|- ✶ASIRM
RP ASIS
Γ|✶ Γ|- Γ|- Γ|-
✶TSA✶ SMHA✶ ●SMNA ●SMKA
Γ|✶
✶ TSA●
Chart showing field marks on issues struck by each mint during the third series, beginning of 351 to
September 352.
Notes: The third series was not struck at the four Gallic mints as they were under Magnentius’ control
throughout the entire time that the third series was struck. They, therefore, only struck the unique Magnen-
tian coin types during this time period.
While both Rome and Aquileia were liberated by Constantius shortly before the end of this series, only Rome
struck a brief issue of FTR-FH coins in this third series, before the fourth series was introduced.
The FTR-FH type is known from Sirmium with ASIRM in the exergue instead of ✶ASIRM. This mark is common
for the HOC SIGNO VICTOR ERIS and CONCORDIA MILITVM types of Vetranio that Constantius continued to
strike into 351, but is not listed in RIC-VIII for the FTR-FH type. It is unclear if it is an official issue or an error.
Examples of FTR-FH from the

third series. Left: Issue from
Siscia (III|- BSIS). Right: Issue
from Antioch, with exergual
mark of the eleventh officina
(Γ|- ANAI).
28
Bastein notes a 4% drop in the weight of this series. Bastien (1985), 92.
124
The Fourth series: September 352 to November 353
A fourth series was introduced when Constantius reduced the weight standard, from 1/60 L to 1/72
L (4.50 g). Many dates appear in numismatic literature for this reduction to 1/72 L, ranging from
mid-351 to late 352. However, the date can be narrowed with some precision, due to the fact that
Constantius’ 1/72 L standard almost certainly begins with the FTR-FH issues at Aquileia and Siscia
with the reverse field mark LXXII, the Roman numeral for 72.
Coins with LXXII were struck at Aquileia shortly after Constantius re-captured the city in
early September 352. There was a brief issue at Aquileia of the very rare GLORIA ROMANORVM
Emperor raising kneeling woman type, followed quickly by the FTR-FH issue with LXXII. Thus we
can date the start of the LXXII issue at Aquileia to sometime in mid-to-late September 352. An
issue of FTR-FH coins with LXXII was also struck at Siscia. Constantius’ re-capture of Siscia in late
September 351 was first followed by an issue of the unreduced (1/60 L) third series FTR-FH type,
bearing the III mark in the reverse field. While we do not know for certain how long this issues
lasted, there is no reason to doubt, given the proximity of the two mints, that the LXXII issue was
introduced at Siscia at the same time as at Aquileia.
In addition to the LXXII mark just discussed, this series saw an intriguing range of different
marks, primarily Δ, S, and E. The Δ was used on the obverse at all of Constantius’ mints from
Sirmium to Alexandria. As the Greek numeral four, it clearly signifies the fourth series, following
the third, which had been marked by a Γ. It is not certain, however, why the series mark was moved
from the reverse for the third series to the obverse for the fourth.
The two other marks used on this series (E and S) are more difficult to interpret. While it is
very tempting to see them as the Greek numerals signifying a fifth (E) and sixth (S) series, there are
several problems with this theory that become apparent when examining the chart below. Unlike
the Δ, the E is not as common or widespread as would be expected of a new series. Furthermore,
despite the sequencing that might be expected from Greek letters (A, B, Γ, Δ, E, S), the E issues fit
better after the S issues than before.
The treatment of Constantius’ Eastern mints in RIC-VIII is somewhat problematic. The
existence of issues at Sirmium, Heraclea, and Nicomedia with both the Δ on the obverse and the S on
the reverse, shows that these two marks, otherwise used individually, must have been consecutive.
Kent therefore placed the issue with the E mark after the S mark at Nicomedia. However, at
Constantinople, Cyzicus, and Antioch, he placed the issue with E before the issue with S. I have
reversed this order at these three mints and placed the E after the S despite the apparent Greek
letter sequencing. Not only does this better match the issue with Δ and S at Sirmium, Heraclea, and
Nicomedia, it also better matches the evidence at Sirmium, Thessalonica, Heraclea, Nicomedia, and
Alexandria where there are Δ and S issues, but no E issues. It appears more likely that these mints
had two issues in sequence, Δ and S, followed by a gap in production, instead of a Δ issue, then a
gap, and then an S issue.
In addition, the E mark is also found on several issues of FTR-FH coins struck at the lighter
1/120 L weight standard introduced in the fifth series (see below). So while some authors have
assigned new series for the coins with Δ, E, and S (as the fourth to sixth series, respectively), I have
decided, based on the difficulty posed by the E, and by the frequent appearance of Δ and S together
on the same coins, to simply assign all 1/72 L coins to one series, the fourth.
125
SHAWN CAZA
-|- -|✶ -|- -|- B on A on Obv, A on Obv, Δ on Δ on Δ on Δ on Δ on Δ on Δ on Δ on
AMB TRP FPLG PARL Obv, LXXII|- LXXII|- Obv, Obv, Obv, Obv, Obv, Obv, Obv, Obv,
Γ|- AQP ✶ ASISu -|- -|- -|- Γ●|- -|- -|- -|- -|-
R●S●P ASIRM TSA SMHA CONSA SMNA SMKA ANA ALEA
A on A on B on A on Obv, A on Obv, Δ on Δ on Δ on Δ on Δ on ●S●|- -|- ●S●|-
Obv, Obv, Obv, LXXII|- -|- Obv, obv, Obv, Obv, Obv, SMKA ANA ALEA
-|- -|- Γ|- AQP● ✶ASISu -✶- II|- S|- A●|- ●S●|-
AMB TRP RUP ASIRM ✶ TSA SMHA CONSA SMNA
A on Δ on A on Obv, A on Obv, Δ on ●S●|- Δ on ●S●|- E on S|-
Obv, Obv, LXXII|S -|- Obv, TSA Obv, SMNA Obv, ANA
-|- Γ|- AQP ●ASISu -L- B●|- -|-
TRPu RUP ASIRM CONSA SMKA
Δ on A on Obv, A on Obv, Δ on Δ on E|-
Obv, LXXII|☧ -|- Obv, Obv, ANA
Γ|- AQP ●ASIS -A- B✶ |-
RFP palm ASIRM CONSA
S|- A on Obv, A on Obv, Δ on ●S●|✶

RP LXXII|U II|- Obv, CONSA
AQP ●ASIS -✶-
palm -A-
ASIRM
S|- A on Obv, Δ on E|-
R✶P II|- Obv, CONSA
●ASIS● ●S●|-
-A-
ASIRM
A on Obv, Δ on
I|- Obv,
ASIS●S● -●S●-
ASIRM
-●S●-
ASIRM
Chart showing field marks on issues struck by each mint during the fourth series,
September 352 to November 353.
Notes: In the chart, the U represents a wreath, while the u represents a small crescent.
An alternative option that has been raised suggests that while the reverse marks are issue marks,
the marks on the obverse were value marks.29 This theory is problematic for several reasons. The
obverse marks were not universal enough to ensure that differing values would be understood.
Furthermore, several different obverse marks were used by different mints at around the same
time. For example, FTR-FH coins of the largest size (series one to three, struck at 1/60 L) appeared
with no obverse mark, then an A on the Western mints, and then no mark once again. Why would
a mark of value appear only on the second series and not the first or third? There is no indication
that the value was changed at this point. Did it need re-confirmation for some reason, and if so,
how could the A help? Coins of the fourth series have as obverse marks an A at Ambianum, Trier,
Aquileia, and Siscia; a B, then Δ, and then nothing at Rome; and a Δ followed by nothing in the East.
If the Δ were intended to signal value, then it lasted for very little time.
The Western mints were under Magnentius’ control when Constantius introduced the fourth
series in the Balkans and the East. Magnentius introduced the weight reduction to 1/72 L in his
29
For example, Kent (1981), 62; and Depeyrot (1992), 65.
126
own mints at about the same time as Constantius, though he did not use any special field marks on
his coins to indicate this. Constantius re-introduced the FTR-FH type into the West as he re-took
control of mints. This occurred when he took Italy from Magnentius in September 352, and again
when he defeated Magnentius and Decentius in Gaul in August 353.
The fourth series continued to be struck in the West for a few months after Constantius
finally defeated Magnentius in August 353, likely until November 353. During this September-to-
November period, these Western mints struck coins of the fourth series, often with an A on the
obverse in what was likely a mistaken continuation of its use from the issues of the second series
struck before they had discontinued this type upon Magnentius’ usurpation.
While 1/72 L is theoretically 4.50 g, coins of this series appear to average 4.3 g.30 The PRD
declined slightly to 20-21 mm. The fineness level fell again to approximately 0.45% (likely meant to
be 1 1/2 scrupulae, or 0.53%).31 This was the last series of FTR-FH to be regularly silvered.
Examples of FTR-FH from the fourth series. Top left: Fourth issue from Aquileia (LXXII|☧ AQP).
Top centre: First issue from Constantinople (Γ|● CONSΓ). Top right: Fifth issue from Constanti-
nople (●S●|✶ CONSΓ). Bottom left: Fourth issue from Antioch (E|- ANH). Bottom right: Sixth
issue from Sirmium (Δ on obverse, ●S●|A BSIRM).
30
Based on 10 examples from Bastien (1985), p.96, and 77 examples personally weighed from private collections. Bastien’s
10 examples yielded a 4.356 g average; my 77 yielded a 4.298 g average and a 4.34 g median. There is, sadly, insufficient
published data to determine whether the average weight dropped from the earlier (Δ) to the later (S) issues within the
fourth series.
31
Bastien (1985), 111.
127
SHAWN CAZA
The Fifth series: November 353 to 356
The fifth series was marked by another weight reduction, this one to 1/120 L (2.70 g). This new
series was not introduced until after a couple of issues of coins of the fourth series had been
struck at Trier after Magnentius’ defeat in August 353. The coinage reform is likely linked to the
celebration of Constantius’ tricennalia (thirtieth year of rule). These celebrations began at Arelate
on 8 November 353 and were accompanied by the widespread striking of gold and silver coins.32 It
is therefore probable that the new series of bronze coins was tied to this same event.
This series can be found with a wide variety of reverse field marks. These marks, however,
were all fairly limited in their usage, and there is clearly no regular field mark that signified the
fifth series. In fact, the majority of coins struck during this series have no field mark at all. This is
interesting, given that of all the series of FTR-FH coins, the introduction of the fifth series was the
most noticeable in terms of size change. The fact that they are the most noticeable by eye might
have meant that there was no need to provide this series with a special identifying mark.
As noted in the discussion of the fourth series, an E appears during the fifth series on one issue
at Arelate, Sirmium, and Heraclea, and on two issues at Constantinople. However, issues with an E
represent no more than a third of the coins struck in this series at Sirmium and Constantinople, and
only a small fraction at Arelate and Heraclea. Therefore, as noted under discussion of the fourth
series, it is unlikely that the E was intended as a universal mark of any series.
The handling of this series in LRBC and RIC-VIII is problematic as both works split this coinage
in two in order to fit it into their structure, which is based on the reign periods of different rulers.
As noted above, there is no reason to believe that the introduction of a new series coincided with
changes in the ruling house just because certain individuals appear on the coinage during that
particular series. In reality, the fifth series of FTR-FH coins began at a point well into the joint reign
of Constantius and Gallus, continued through Constantius’ period of sole reign, and into the early
part of the joint reign of Constantius and Julian.
In numerous cases, both LRBC and RIC-VIII provide separate catalogue numbers for what is in fact
the same coin issue. For example, at Cyzicus, the FTR-FH with the exergual mark SMKA, and no field
mark, is catalogued as LRBC-II-2496 for the 351 to 354 period (the joint reign of Constantius and Gallus)
and as LRBC-II-2498 for the 6 November 355 to 3 November 356 period (the joint reign of Constantius
and Julian). The same coin is catalogued in RIC-VIII as RIC-104 and RIC-111 for these two periods,
respectively. Both works place Constantius’ sole reign period with his joint reign with Gallus. Despite
the use of separate numbers, these coins are identical in field and exergual marks, diameter, and weight.
This situation is repeated during this series across most mints. This frequent provision of two catalogue
numbers for the same coin has created a fair amount of confusion. It also obscures the fact that the start
and end of coinage series, like this one, are not necessarily linked to changes in the ruling house.
Coins of the fifth series averaged 2.48 g, lower than the theoretical 2.70 g.33 The coins of the
fifth series are noticeably smaller than those of the preceding series. The PRD declined substantially
to 17-17.5 mm. They appear to have a slightly higher fineness level than those of the fourth series --
approximately 0.7% (2 scrupulae). This is not as odd as it may at first seem. Given the weight reduction,
the 1/120 L coins still had, on average, less silver (18.9 mg) than those struck at 1/72 L (20.25 mg).
32
Bastien (1985), 70.
33
Bastien (1985), 96, and Depeyrot (1992), 66, taken from 679 examples.
128
-|- -|- -|- -|- -|- -|- -|- A|- -|- -|- -|- -|- -|- -|-
TRP FPLG PCON RP AQP ASIS ASIRM SMTS SMHA CONSA SMNA SMKA ANA ALEA
-|- -|- D|- -|- -|- -|- -Ev- -|- E|- ●|- A|- ✶ |-
TRPu CPLG PCON R✶P ●AQP● ASISz ASIRM SMTSA SMHA CONSA SMNA SMKA
E|- -|- II|- -|- -|- X●|-

PCON RP palm AQP● ASISD ASIRM● CONSA
-|- II|- E●|-
R palm P AQP✶ CONSA
-|- II|- E|-
RUP AQP CONSA
II|- C|A
AQP CONS
branch
II|✶ -|-
AQP● CONSA●
II|✶
AQP
branch
Chart showing field marks on issues struck by each mint during the fifth series, November 353 to 356.
Notes: In the chart, the U represents a wreath, while the u represents a small crescent. The z in this chart
represents a symbol on the coin that resembles a backwards z and likely represents something other than a
letter, possibly a thunderbolt. Julian appears on all of the issues below the thick line.
The second issue at Lyons has a G in the exergue according to LRBC and RIC-VIII. Bastien, however, records this
as a C, which appears to be a more accurate reflection of what is actually found on the majority of examples.
The first issue at Thessalonica uses the officina letter as a field mark. Thus it appears as A|-, B|-, Γ|-, Δ|- and E|-.
At Arelate, Ferrando placed the issue E|- in the sixth series, after the M|- issue based on his observation that
they were smaller (16 mm diameter vs. 19 mm).34 While this order is possible, it would go against the entire
pattern of the sixth series, all of which used an M, with the exception of one issue at Siscia with an L (see
below). It would also seem more natural for the E to follow the D than the M. The apparent decrease in
diameter has other possible explanations, including statistical anomalies stemming from the small sample
size, or the possibility that the mint was trying to conserve metal during that particular issue and scrimped
on the quantity used. There are no indications of any official decrease of the weight standard, or die size,
within either the fifth or sixth issues. I have therefore placed this E|- issue during the fifth series.
RIC-VIII also lists examples from Nicomedia with ✶|-, apparently based on a unique specimen. This is possibly
a misreading of a coin from Cyzicus, as coins from these two mints are easily confused on worn examples.
The rise in fineness in this, and the next, series to 0.7% was clearly intentional. The fact that the
fineness level quickly dropped to 0.1% for the following SPES REIPVBLICE type, indicates that when
the Roman state wanted to eliminate the silver content of a coin type, it could do so quickly and
accurately. Had it therefore been intended that the fifth and/or sixth series had no silver content,
we could expect the fineness levels to be closer to 0.1% than 0.7%.
The FTR-FH coins of the fifth and sixth series do not appear to have been silvered. As these
were the last Roman bronze coins to contain intentionally added silver (with the brief exception
of Julian and Jovian’s large AE1 denomination of 362 to 364), it is possible that the practice of using
silvering to signal that the coin had a silver content was simply abandoned shortly before the
addition of silver itself was abandoned. Overall, the late FTR-FH issues could more properly be
34
Ferrando (2010), 34.
129
SHAWN CAZA
considered leaded-copper than bronze. Tin levels were as low as 1-2 %, while the percentage of lead
had risen to levels up to the mid-20s.35
Examples of FTR-FH from the fifth series. Left: Second issue from Lyons (-|- CSLG). Centre:
Seventh issue from Aquileia (II|✶ AQS●). Right: First issue from Thessalonica, with field mark of
the fifth officina (E|- SMTS).
The Sixth series: mid-356 to November 358
A further weight reduction occurred in 356, with the introduction of the sixth series. Unfortunately,
it has not proved possible to date this reduction with any greater precision. Coins of the fifth series
are quite common for Gallus (who ruled until circa October 354), but are scarcer for Julian, who
became Caesar in November 355.36 We can therefore almost certainly date the introduction of the
sixth series to 356, though whether in January, or later in the year, remains uncertain. The fact that
Aquileia had time to strike four issues of the fifth series that included Julian, and were thus struck
from November 355, implies the start date of the sixth series was perhaps closer to the middle than
the start of 356.37
The sixth series was distinguished by the mark M. This mark appeared in the reverse left
field of almost every issue during this series. There were, however, a few exceptions. At Rome the
M was instead found as part of the exergual mint mark. At Lyons and Aquileia, the M was found on
the obverse, though it appears in the reverse field as well on several of these issues at Aquileia. A
single issue at Siscia used an L in the reverse field instead of an M.
The exact meaning of this M is uncertain. The view of Callu, also put forth by Kent, that
the M of the sixth series meant that the bronze coin was worth 1/1,000th of a gold solidus is
35
Bastein (1985), 112, and Brickstock (1987), 59.
36
While the date of Gallus’ arrest at Poetovio (modern Ptuj, Slovenia) and subsequent execution near Pola (modern Pula,
Croatia) are unknown, we do know that Gallus departed Antioch in early September 354, and dallied for games in
Constantinople. He was therefore likely arrested in late October or November 354 and executed before the end of the
year. Crawford (2016), 99.
37
The historical sources do not detail any political issues of this time that might have led to a change in the coinage.
Instead, they are concentrated on Constantius’ religious struggles, especially with Athanasius, and on Julian’s campaigns
against the Germanic tribes along the Rhine.
130
not correct.38 There is no proof that the bronze coinage was tied to the gold in this manner at
this time. Furthermore, a value of 1/1,000th of a solidus would make the FTR-FH coin grossly
overvalued. According to the Codex Theodosianus XI 21.2, gold was worth 1,800 times more than
bronze by weight, and the bronze coins were lighter than the gold solidus. Looking at the value of a
pound of gold, Depeyrot has shown that by time of the fifth FTR series, one gold solidus was worth
approximately 2,880 FTR-FH coins.39 The M on the subsequent series must, therefore, have had a
different meaning.
Several authors believe that the FTR-FH coin was worth 1,000 denarii communes (d.c.) from
its introduction in 348, and that the M on the sixth series must be the Roman numeral for 1,000,
representing a value mark of 1,000 d.c. Taking into account weight and fineness levels, the large
denomination (including the FTR-FH) of 348 had an intrinsic value approximately nine times that
of the centenionalis of 341 to 348 AD.40 As this centenionalis is widely believed to have been worth
100 d.c. since its introduction in the 310s, the new maiorina coin must have had a value of at least
900 d.c., or the Roman state would have been on the losing end with its introduction.41 This makes
a value of 1,000 d.c. for the FTR-FH at its introduction in 348 likely.
The subsequent reductions in weight and fineness, and therefore in intrinsic value, of the
FTR-FH coin would account for some of the inflation of the time. However, a recent paper by
Bagnall and Bransbourg seeks to explain the significant increase in prices during the mid-fourth
century through a major increase in value of the bronze coinages from 1,000 d.c. to 10,000 d.c.
sometime after 353.42 This idea, with some modification, could explain the M that is found on
almost every coin of this sixth series.
Could this M stand for a myriad, in other words, 10,000 d.c.? Bagnall and Bransbourg place
the value increase to 10,000 d.c. at the time of the reduction of the FTR-FH coinage to ‘AE3’; in other
words, with the introduction of the fifth series, which I have dated to November 353. However, the
papyri on which the price increases are based on are not dated with clear certainty.
The fifth series represented a significant size and weight reduction from 1/72 L (4.5 g) to
1/120 L (2.7 g). This reduction alone, without any re-valuation, would have yielded significant
profit to the state - and led to further inflation. By contrast, the sixth series saw no major change
in size (the die diameter dropped by about 1 mm), weight (dropped from 2.7 g to 2.25 g), or fineness
(which did not change). Why, therefore, was the M found on almost every coin of this series? To
the normal user the fifth and sixth series looked almost identical. The most likely event that would
require a universal marking on the new series was a change in value. If the value had been 1,000
d.c. since 348, then the M must stand for something else. The introduction of the myriad (or 10,000
d.c.) value for the sixth series would require only moving Bagnall and Bransbourg’s date a little over
two years later, and would better match the numismatic evidence.
The M also appears on an extremely rare bronze type struck during this series, with a reverse
design consisting entirely of a large M. This type, struck at Rome, was obviously related to the M
field mark and was likely struck at the beginning of this series.
38
See Kent (1981), 65 and Depeyrot (1992), 72.
39
Depeyrot (1992), 72.
40
Cope (1974), 235 - 236.
41
Depeyrot (1992), 72.
42
Bagnall and Bransbourg (2019).
131
SHAWN CAZA
Though clearly lighter than the fifth series, it is not absolutely certain what the new weight
standard of the sixth series was. While Kent thought that it represented an official weight reduction
to 1/137 L (2.36 g), Bastien disagreed, citing both the level of reduction in the weight averages, and
the fact that 1/137 L was an unwieldy non-duodecimal fraction.43 He suggested a 1/144 L standard
(2.25 g).
Coins of the sixth series average 2.25 g, which appears to indicate that 1/144 L is too low
when taking into account the need for some production loss and the fact that average weights are
almost always a bit lower than the theoretical standard.44 However, as Bastien noted, the 1/137 L
option is unlikely as it is a very difficult fraction to manufacture. The 1/144 L option is much easier
and, perhaps more importantly, represents half of the earlier 1/72 L standard. Barring further data,
I have chosen to follow Bastien’s 1/144 L standard.
The two scrupulae fineness level (0.7%) introduced for the fifth series appears to have lasted
until the end of the FTR-FH coinage. The sixth series has an average fineness near this level (0.65%),
though their lower weight standard meant they had less silver (15.6 mg) than the fifth series. The
PRD declined slightly from 17-17.5 mm to 16-16.5 mm.
M in M|- -|- M in Obv, M|- M|- M|- M|- ●M●|- M|- ●M●|- M|- M|-
Obv, PCON R●M●P -|- ASISR or ASIRM SMTSA SMHA CONSA SMNA SMKA ANA ALEA
-|- ●AQP ASISM
CPLG
M in -|- M in Obv, M|- M|- M|- ●M●|-
Obv, RMP✶ -|- ASISz ASIRM● SMHA● CONSA●
-|- AQP
MPLG branch
M in Obv, M|- M|- ●M●|-
-|- ASISD ●ASIRM● CONSA✶
●AQP
branch
M in Obv, M|● M|- ●M●|-
-●- ASISD ASIRM✶ CONSA
AQP palm
M in Obv, M|- M|- ●M●|:
-✶- ASIS ●ASIRM✶ CONSA
●AQP●
M in Obv, M|-
M|- ASISL
AQP
M in Obv, L|-
M|- ASIS
✶ AQP
branch
Chart showing field marks on issues struck by each mint during the sixth series, 356 to November 358.
Note: The z in this chart represents a symbol on the coin that resembles a backwards z and likely represents
something other than a letter, possibly a thunderbolt.
The first issue at Lyons has a G in the exergue according to LRBC and RIC-VIII. Bastien, however, records this
as a C, which appears to be a more accurate reflection of what is actually found on the majority of examples.
43
Bastien (1985), 93 and 96.
44
Depeyrot (1992), 66, taken from 103 examples (2.26 g); and Alram and Schmidt-Dick (2007), section 2.4.14, taken from
3,474 examples (2.25 g).
132
An issue in RIC-VIII from Siscia with a wreath in the exergue appears to be based on only two examples. Kent
notes that one requires confirmation. The other might be an error for a D, which was a common mark. It is
therefore not listed here.
Though at first glance the Siscia issue with L|- appears to break the otherwise universal rule of the sixth series
having an M somewhere on the coin, the fact that an L was included in the exergue of the previous issue (M|-,
ASISL) makes it likely that it was indeed part of the sixth series.
Examples of FTR-FH from the sixth series. Left: Seventh issue from Siscia (L|- ΔSIS). Centre:
Third issue from Sirmium (M|- ●ASIRM●). Right: First (and only) issue from Thessalonica
(M|- SMTSE).
THE END OF THE FALLING HORSEMAN TYPE
In November 358, the FTR-FH was replaced by a new, and smaller, type, SPES REIPVBLICE Emperor
standing holding spear and globe. This change not only ended the FTR coinage, but changed the main
message of the bronze coinage from overtly military designs accompanied by the message ‘the
return of fortunate times’, to the image of a lone helmeted figure holding an orb (almost certainly
either the Emperor or the Caesar) accompanied by the message ‘the hope of the Empire’.
The dating, and reason, for the end of the FEL TEMP REPARATIO coinage and the introduction
of the new SPES REIPVBLICE type is not absolutely certain. LRBC does not specifically date the
new type, but places its introduction in the 355 to 360 period, along with the last two FTR-FH
series. Similarly, RIC-VIII does not provide a specific date, though Kent notes that it must have been
introduced ‘appreciably before’ 360, given the large quantity of SPES REIPVBLICE coins known for
Julian as Caesar. Bastien and Depeyrot, however, dated it to 358.45
Callu went further. He believed that the issue was intended to mark the celebrations of
Constantius’ 35th anniversary.46 The seven-year-old Constantius had been named Caesar on
8 November 324, so celebration of his 35th anniversary of reign would normally have begun on
8 November 358. This dating would appear to fit the introduction of the SPES REIPVBLICE type
perfectly. While Kent noted in RIC-VIII that Constantius may have begun celebrations of his 35th
anniversary earlier, during his visit to Rome in April to May 357, Ammianus Marcellinus does
45
Kent (1981), 40; Bastien (1985), 93; Depeyrot (1992), 67.
46
Callu (1986), 210.
133
SHAWN CAZA
not mention the anniversary, simply Constantius’ desire to visit Rome and celebrate his recent
victories.47 It does not make sense to date the start of the SPES REIPVBLICE coinage as early as April
357, which would be too early to accommodate the huge numbers and issues of the sixth series
of FTR-FH coinage. It is more likely that the type was introduced in November 358 to mark the
Empire-wide celebrations of his 35th anniversary.
The six FEL TEMP REPARATIO Falling Horseman Series Summary
Series Weight Average Marks Issue Struck For

Standard Weight Period
1st 1/60 L 5.3 g None on obverse April 348 to mid-349 Constantius II, Constans
2nd 1/60 L 5.2 g A on obverse Mid-349 to end 350 Constantius II,
Constans, Magnentius
3rd 1/60 L 4.7 g III or Γ on the reverse Beginning 351 to Constantius II, Gallus
September 352
4th 1/72 L 4.34 g Often Δ on the obverse, September 352 to Constantius II, Gallus
LXXII, S or E on the November 353
reverse
5th 1/120 L 2.5 g Wide variety, though November 353 to Constantius II,
often none, on reverse. 356 Gallus, Julian
6th 1/144 L 2.25 g M, usually on reverse Mid-356 to Constantius II, Julian
November 358
THE DESIGN OF THE FEL TEMP REPARATIO FALLING HORSEMAN TYPE
This design of a large Roman footman spearing a fallen enemy horseman was unique in Roman
coinage. It was far more common on Roman coins to show a large Roman horseman spearing a
small enemy footman. In addition, many Roman gravestones show a Roman cavalryman spearing
an infantryman, but few, if any, show the reverse situation.48
Who is this large footman figure on the FTR-FH type -- the spearman stabbing the falling
horseman? In his 1958 work, Kraft believed that the large figure on all of the FTR types was
intended to represent the Emperor. On several types (Galley, Emperor & Captive, and Emperor with Two
Captives), the figure, though in military garb, holds a labarum and clearly wears a diadem. These
two Imperial symbols -- labarum and diadem -- mean that these figures can represent none other
than an Emperor.49
The large figure on the FTR-FH type has generated more discussion. This large figure has
been described in different sources as ‘soldier’ (Stevenson 1889, RIC-VIII and Abdy in Metcalf 2012),
47
Kent (1981), 14, and Ammianus Marcellinus, XVI, 10 (Text from LacusCurtius site: http://penelope.uchicago.edu/
Thayer/E/Roman/Texts/Ammian/16*.html ).
48
Kraft (1958), 154. Failmezger (1992) claims this image can be found on a first-century sarcophagus. However, I have
not been able to find a clear and close parallel of this design amongst the melee portrayed on such sarcophagi.
49
Kraft (1958), 143.
134
‘warrior’ (Mattingly 1933), ‘Emperor’ (Bastien 1985), and even ‘Virtus’ (LRBC). He is helmeted,
dressed in military garb, and holds a spear instead of a labarum. Though it is less clear cut than
for the diademed figures on the other types, I believe that this large figure must also represent
an Emperor. Kraft has noted that the design of the outfit is not that of a common soldier but
is more applicable to a high ranking official.50 Furthermore, other than the helmet, the outfit is
generally the same as that worn by the Emperor figure on the other FTR types. Kraft also noted
that the figure cannot be Virtus, who was almost always portrayed as an armoured female figure.
In addition, if all of the other types in the FTR series clearly show an Emperor, it is likely that the
same was intended on the FTR-FH type as well. All of these factors argue strongly for the large
figure being the Emperor as well.
In 1933, Mattingly wrote that the FTR-FH type might have represented the famed ‘Achilles
stabbing a Trojan’ motif, given that the Trojans were often portrayed as mounted warriors in
Roman-era iconography. However, as Kraft has pointed out, the Romans traced their ancestry to
the Trojans, and not the Greeks, and it is therefore highly unlikely that they would choose a design
highlighting the slaughter of their ancestors.51
Instead, Kraft believed that the type was intended to portray Constantius striking down a
Persian prince in battle and believed that it refers to Constantius’ victory over the Persians at the
Battle of Singara in 344, when the Sassanian Crown Prince was killed. However, the year, and the fate
of the Persian prince, are both entirely uncertain given the confusing historical documentation.52
The 344 battle of Singara is not the only possible incident that might be portrayed. Constantius
was made Caesar, at age 7, on 13 November 324. His first significant military or political action came
in 335 when he was sent by Constantine I, under the tutelage of the Praetorian Prefect of the East,
Ablabius, to take command of the defences against the Persians. Soon after Constantius’ arrival at
Antioch, the Persians invaded Roman territory. The Roman army, ostensibly under the command
of Constantius, defeated a large Persian army at Narasara, near Amida. The Persian general Narses,
the younger brother of the Sassanid Persian King of Kings Shapur II, was killed during the battle.
Thus Constantius was first associated in the minds of many Roman citizens with a military victory
over the Persians and the killing of a Persian prince.53
It is also of course possible that the design was intended to be generic, and to simply signify
Constantius’ puissance over the Persian enemy instead of a specific event of 335 or 344. In any
event, the horseman on the FTR-FH type wears a Phrygian cap and trousers, features which, in the
language of Roman imagery, indicate a Persian.
The image of a large figure leaning over the flank of an animal and striking a death blow with
a spear may also have reminded many viewers of the Mithraic tauroctony (bull sacrifice) scene.54
Mithraism had been immensely popular among Roman troops since the third century, and though
the Roman Empire was formally Christian by the time the FTR-FH type was struck, worship of
50
Kraft (1958), 144.
51
Kraft (1958), 147.
52
Kraft (1958), 158 - 159; Crawford (2016), 54 - 57.
53
Crawford (2016), 50-51.
54
While this was my own observation based on the design in question, I have since discovered that the same point was
made by Moreno Sagramora in his book on the coinage of Aquileia. Sagramora (2010), 284.
135
SHAWN CAZA
Mithras remained.55 This is not to say that the scene on the coin was intended in any way to refer
to Mithraism, just that its style had similarities that would have been recognized by at least some
of the population.
Nineteenth-century Engraving of a Mithraic Tauroctony relief.
FEL TEMP REPARATIO FALLING HORSEMAN SUB-TYPES AND VARIATIONS
There is significant variation in the design of the FTR-FH type. The position of the Emperor, the
horse and the horseman all vary, as does the horseman’s headgear and clothing.
Sub-types based on the position of the horseman
The position of the horseman is traditionally used to divide the FTR-FH type into sub-types. In
LRBC, Carson and Kent identified four sub-types based on the position of the horseman vis-à-vis the
horse. Their four sub-types are:56
FH1 Horseman in kneeling position beside horse

FH2 Horseman sitting on ground beside horse
FH3 Horseman falling from horse and raising arm behind
FH4 Horseman falling and clutching horse’s neck
However, in 1981 in RIC-VIII, Kent reduced the number of sub-types to three as he believed that
the previous FH1 variation was simply a different form of FH4 wherein the horseman’s knees were
bent. Kent’s three sub-types are:57
55
Mithraism in fact continued among some elements of the Roman military into the late fourth century. The last known
Mithraic inscription in Rome dates to 391. It might have lasted even longer in the provinces. Beard et al. (1998), vol.1, 387.
56
Hill, Kent and Carson (1978), 108.
57
Kent (1981), 38.
136
FH1 Horseman sitting beside fallen horse, facing his attacker, to whom he lifts his arms
FH2 Horseman lying along the back of fallen horse, arms apparently clutching its neck
FH3 Horseman clutching horse’s neck with right hand, head turned back towards
attacker, to whom he raises his left arm
In his 2002 book on Roman bronze coins, Failmezger used LRBC’s four sub-types instead of adopting
Kent’s more recent three. He noted that LRBC’s FH1 and FH4 can be distinguished by the direction
that the horse’s head faces. Despite this difference, I believe that Kent was correct in grouping
LRBC’s FH1 and FH4 together into one sub-type. As is discussed below, the main diagnostic feature
for these two variations is the fact that the horseman leans forwards, away from the Emperor and
towards the left side of the coin.
While I believe that Kent’s breakdown into three sub-types is best, I will use capital letters
instead of numbers to avoid confusion between the different systems.
Sub-Type Horseman Design LRBC RIC-VIII

A Horseman facing Emperor 2 1
B Horseman leaning forward 1 and 4 2
C Horseman reaching back 3 3
Sub-type A) Horseman facing Emperor
On this sub-type the horseman’s body faces the Emperor with

one or both arms usually raised in apparent supplication. The
horseman is beside the horse (on its left side) and is either sitting,
with legs stretched out before him, or kneeling, with legs bent.
This sub-type was struck only in the West and was thus a
product of Constans’ mints. However, as with the entire FTR-FH
type, it was mainly struck for Constantius II. It was also struck for
Constans, though perhaps not at every mint, and was struck later
for Constantius Gallus at Thessalonica.
While the sub-type is early, it was not the first one struck.
It was used during the second series (mid-349 until 350) at all Western mints except Rome. It
followed sub-type B, which had been used for the first series (April 348 to mid-349), except at Trier,
where the FTR-FH type was not struck in the first series. In most cases it was in turn followed by
sub-type C. It was also struck by Magnentius for Constantius in 350 and 351 at Arelate, Rome, and
Aquileia.
Though sub-type A ceased to be struck at most mints by the end of 350, it was struck until
late February 351 at Rome and Aquileia, and until at least late March 351 at Thessalonica, given
the existence of some coins for Gallus. It was then struck again briefly in late 353 at Trier and
Lugdunum after the defeat of Magnentius, at the first reduced weight standard (1/72 L). All other
examples of this sub-type were struck at the initial weight standard (1/60 L).
137
SHAWN CAZA
Sub-type B) Horseman leaning forward
On this sub-type, the horseman’s body leans forward, away from

the Emperor. The horseman is either portrayed seated on the
horse, in which case he leans forward over the horse’s neck, which
he grasps with both hands, or is kneeling or lying beside the horse.
When the horseman is seated on the horse, the horse’s head is
tucked under, towards the right, and it appears as if the horseman
is about to fly forward over the horse’s head. By contrast, when
the horseman is kneeling or lying on the ground, the horse’s head
is usually extended forward, towards the left.
This sub-type was the first sub-type struck. It was struck
during the first series (April 348 to mid-349) at all active mints.
At several of the Western mints it was struck only for the first series (April 348 to mid-349)
and was replaced in the second series (beginning mid-349) with sub-type A. It was also struck by
Magnentius for himself at Arelate early in his reign. However, it was used at Constantinople and
Antioch right through to the end of the FTR-FH series, though not as the sole sub-type. It is thus
found in all sizes in the East.
Sub-type C) Horseman reaching back
On this sub-type the horseman’s body faces forward, but he turns

to reach back towards the Emperor. He is usually seated on the
horse, though sometimes the horseman sits or kneels beside the
horse, in which case he often grasps it with his other arm. The
horseman usually has his left arm raised, though it can be the
right arm, or more rarely both arms. His other arm often grasps
the horse’s neck, though occasionally it hangs limply down. The
key design feature is that at least one arm always reaches back to
the Emperor.
This sub-type first appeared in mid-349 at Constantinople,
Nicomedia, and Antioch and was thus the last sub-type to appear.
It was a product of Constantius’ mints. It soon spreads to all other mints and is by far the most
common sub-type.
In the West, sub-type C appears after the liberation of mints from Magnentius, sometimes, as
at Trier and Lugdunum, after a brief issue of sub-type A. It is then the sole sub-type for the rest of
the FTR-FH issue. At Thessalonica, it appears only with the fourth series in September 352, and at
Rome with the fifth series in November 353.
In the East it appears in the second series in mid-349, or the third series in the beginning of
351. It was struck at all of these mints until the end of the FTR-FH type, though not always as the
sole sub-type.
138
VARIATIONS BASED ON THE DESIGN OF THE HORSEMAN
Different designs for the horseman were clearly used by different mints at different times, though
to date there has been little study of the details. Mattingly, LRBC, and Failmezger did not break the
coinage down by horseman type. Guido Bruck examined only the appearance of the hat and beard
on sub-type C. In RIC-VIII, Kent proposed four designs of horseman, but unfortunately did not
apply this classification in his coin descriptions to every mint.
Kent’s 1981 breakdown of horseman designs is as follows:58
a. Horseman is bare-headed and beardless

b. Horseman is bare-headed and bearded
c. Horseman is wearing Phrygian cap and beardless
d. Horseman is wearing diadem and bearded
Some researchers believe that there are more variations. On her website (www.catbikes.ch/
helvetica/feltemps.htm), Dane Kurth has categorized a greater number of horseman designs,
including six different helmet types and a ‘hair straight up’ design. She also makes the important
observation that the so-called ‘diadem with ties’ design was more likely intended to portray long
braids, a common hairstyle among Eastern nomadic horsemen. Failmezger lists ten different
variations of barbarian depicted across the entire series, though these include many details that
were almost certainly not official variations, such as two hair braids versus three hair braids.59
The challenge lies in determining which design differences had some significance in terms
of issues or mints, and which were simply variations by individual engravers. Furthermore, the
problem of worn or poorly preserved coins, and of the many unofficial imitations of the FTR-FH
type, makes the clear identification of official variations challenging.
While the horsemen can be classified, and perhaps even identified, by their attributes, such
as hats, hairstyle, and clothing, not every design element was relevant for the identification of
issues. Thus the Phrygian cap, which was traditionally a soft cap, could be portrayed with a point
that was long or short, wide or thin, and with a clear bend or almost straight. It could also be
portrayed with a wide rim, a narrow rim, or no rim at all. This leads to dozens of minor variations
in the actual design of the cap, though all are likely intended to represent the same thing -- an
Eastern horseman’s cap. Likewise, the bare head could be shown with hair engraved from side to
side, with hair standing up vertically, or, most commonly, with no hair lines visible at all, whether
due to wear or economy of design.
Overall, I believe that the wide variety of designs can be grouped into three clear horseman
types. Once again, I have introduced a new coding, changing Kent’s lower case letters (a, b, c) to
lower case Roman numerals (i, ii, iii) to avoid confusion.
58
Kent (1981), 39.
59
Failmezger (1992), 18.
139
SHAWN CAZA
Horseman Type Horseman Likely Portrays Kent

i Wearing Phrygian cap, usually wearing Persian or Eastern enemy c
trousers.
ii Bare head, often wearing a tunic, Germanic or Western enemy a and b
sometimes trousers.
iii Beard and braids, usually wearing Nomadic enemy d
trousers.
i. The Phrygian cap design was the most common. It was used during the first series at
almost all mints and remained important throughout the Empire until the end of the FTR-
FH type in 358. It was used for a rare issue for Constantius Gallus at Ambianum. The only
mints it was not used at were Sirmium and Antioch.
ii. The bare-head design is less common than the Phrygian cap design. It began at
Thessalonica in the first series and spread to nearby Constantinople for the second
series.60 It then became very widespread in the West with the first weight reduction (the
fourth series) of September 352. This design can be found with and without a beard. It is
often difficult to tell if a beard is portrayed.
c. The beard and braids design is the rarest, and appears to have been used at only three
mints. It began at Antioch during the first series and was used there until the end of the
FTR-FH coinage in 358. It was adopted at Heraclea for the third and fourth series, and
was used briefly at Ambianum during the very short time the mint struck the FTR-FH
type in late 353. This last fact is interesting given that Ambianum, modern Amiens in
northern France, was located about as far from the range of nomadic horsemen as could
possibly be. This may indicate that an engraver from Antioch or Heraclea began working
at Ambianum when Constantius liberated it.
Overall, we can see that there is no real chronological significance to the three horseman designs.
All three were used from the beginning of the FTR coinage in April 348 and all three were in use
until the end of the FTR-FH type in 358. There is a slight geographic significance for the bare-head
design (ii), which is not found east of Constantinople.
The occurrence of sub-type and horseman design by series and mint can be seen in the
following chart. The smaller the coin (i.e., the latter the series), the more difficult it can be to
distinguish horseman types (i to iii). Combinations in italics and underlined are either very scarce,
or are possibly other combinations that are poorly executed and hard to distinguish.
RIC-VIII mentions only the pointed hat (design i) for this issue at Thessalonica (RIC 114-116). However, the version with
60
bare head (design ii) is numerous and can be found illustrated in Failmezger and on wildwinds.com
140
Series 1. 1/60 L 2. 1/60 L 3. 1/60 L 4. 1/72 L 5. 1/120 L 6. 1/144

4/348- mid-349 - 1/351 - 9/352 - 11/353 - L 356 -
mid-349 12/350 9/352 11/353 356 11/358
Ambianum -- -- -- Bi, Ciii -- --

Treveri -- Ai -- Cii Ci, Cii --
Lugdunum Bi Ai -- Ai Ci, Cii Ci
Arelate Bi Ai -- Ci Ci, Cii Ci, Cii
Roma Bi Ai Bi Bi Ci, Cii Ci, Cii
Aquileia Bi Ai -- Cii Cii Cii
Siscia Bi Ai Cii Cii Ci Ci, Cii
Sirmium -- -- Cii Cii Cii Cii
Thessalonica Bi, Bii Ai Ai Bii, Ci, Cii Ci, Cii Ci, Cii
Heraclea Bi, Bii — Cii, Ciii Cii, Ciii Ci Ci
Constantinople Bi, Ci, Cii — Bi, Cii Bi Bi, Ci, Cii Bi, Ci, Cii
Nicomedia Bi, Bii — Ci Ci, Cii, Ciii Ci, Cii Ci, Cii
Cyzicus Bi — Ci, Cii Ci, Cii Ci Ci
Antioch Biii, Ci — Biii, Ciii Biii, Ciii Biii, Ciii Biii, Ciii
Alexandria Bi, Biii — Ci, Cii, Ciii Ci, Cii Ci, Cii Bii, Ci, Cii
VARIATIONS BASED ON THE POSITION OF THE EMPEROR
In addition to the variations in the position of the horseman and his horse (discussed above), the
position of the Emperor also varies. This variation can be seen primarily in the position of his right
leg. However, it is difficult to divide the design into neat types, as the Emperor is in fact found in a
continuum of positions from standing to kneeling.
On most coins, the Emperor’s right knee is in contact with the rear flank of the horse. On
many coins, especially those struck at Lugdunum, Arelate, Rome, Siscia and Constantinople, the
Emperor’s knee is bent greater than 90 degrees and his lower leg extends back horizontally -- it does
not reach down towards the ground. On other coins, the Emperor’s leg is bent less than 90 degrees
and extends down towards the ground. On some examples, especially those struck at Eastern mints,
the Emperor’s right leg is clearly planted on the ground, or on the horseman’s shield, which lies on
the ground. However, on many coins, including many struck at Arelate, Sirmium, Thessalonica and
Heraclea, the Emperor’s right leg is portrayed between these two positions, bent at approximately
90 degrees with the lower leg pointing towards the ground, but not reaching it.
There are only two truly different design variations. On the first variation, the Emperor’s
right leg is extended up and forward so that his right foot rests on the horse’s back. This is typically
found on some early coins struck at Thessalonica -- the same TSA❋ issue of early- to mid-349 that
uses the globe bust (see above). On the second variation, the Emperor kneels on the horseman’s
back. This is typically found on early coins of sub-type B struck at Siscia.
Overall, the different positions of the Emperor appear to be no more than mint design
variations and do not appear to have any other significance.
141
SHAWN CAZA
The obverse
On the FTR-FH type, Emperors always used a diademed bust. This is almost always the pearl diadem,
draped and cuirassed bust facing right. The rosette diadem is sometimes used instead of the pearl
diadem, though this likely has no significance. The only true bust variation is the special globe
bust (described under The First Series above). The Caesars Constantius Gallus and Julian are always
shown with bare head, draped and cuirassed bust, though there is a variation for Julian with bare
head, cuirassed bust at Rome.
142
IMITATIONS OF THE FEL TEMP REPARATIO FALLING HORSEMAN TYPE
Imitations of the FTR-FH coinage are very common. Imitations of the larger size (the first four series
dating from 348 to 353), tend to be of similar size, weight, and quality to the original coin types.
However, they contain much less silver than the originals. While it is therefore likely that these
imitations, which began to appear shortly after the introduction of the new type, were primarily
made for profit, some were also intended to make up for a shortfall of official coinage.61 Copies of
the FTR-FH type are even known in lead, and such coins could only have been tolerated to make up
for a serious local shortage.62
By far the most common type of imitative in this period was the small FTR-FH type (of the
last two series). These coins were especially prevalent in border regions such as Britain, northern
Gaul, the Balkan limes and the Middle East. However, they were very widespread and are found
all over the Empire. These later FTR-FH imitations represent an attempt to make up for a serious
61
Bastein (1985), 9.
62
Ferrando (2010), 51.
143
SHAWN CAZA
shortage of official coinage. The shortage of bronze coinage during this period was exacerbated
by several factors: the demonetization of earlier coinage, mint closures and the deliberate policy
of Constantius.
A law of the Codex Theodosianus (IX.23.1 of 8 March 354) issued at Arelate after Constantius’
tricennalia celebrations, ordered demonetized bronze coins to be returned to the State. These likely
included the large denominations of 348 to 353, and all Magnentian coinage. This law was intended
to ensure that the State could gain the profits from melting these coins down and re-coining the
metal. In the Western part of the Empire, this resulted in the withdrawal of most of the older
currency.
The shortage was exacerbated by mint closures. Britain had not had an operating mint
since London closed in 325. The short-lived mint of Ambianum, in northern Gaul, was closed by
Constantius not long after his defeat of Magnentius in late 353. The supply once provided by these
two mints was to have been made up by the mint of Trier. However, Trier was out of commission
from 354 to 357 because of the destruction caused by the invading Franks. These factors resulted
in a serious bronze coin shortage in northern Gaul and Britain.
The shortage in the West was also exacerbated by Constantius’ deliberate policy of ‘starving’
first Constantius Gallus, and then especially Julian, of fresh coinage. In these areas, the FTR-FH
imitations continued to be struck and used until 364, when they were quickly superseded by official
Valentinianic coinage. An FTR-FH imitation with Julian as Augustus on the obverse, found in the
hoard from Boulogne-sur-Mer, France, shows that imitations of this type were struck after Julian
was proclaimed Emperor by his troops in February 360, more than a year after the FTR-FH type
ceased to be officially struck.63
It is also possible that the general requirement for bronze coinage across the Empire increased
during this period as the concomitant increase in gold and silver coinage at this time would have
increased the need for bronze divisional coins to serve as small change. This need would have been
especially prevalent in regions with a large military presence and hence a great deal of interaction
between well-paid soldiers and the local inhabitants.64
These imitations were clearly tolerated, as they are frequently found mixed in hoards with
regular coins. Interestingly, many sizes are found mixed together. They were also tolerated by the
Roman military, as demonstrated by the fact that they are found on military sites in significant
numbers. However, they were likely not made by the Roman military. They represent an even
higher proportion of finds on civilian sites.65
These late FTR-FH imitations are often obvious by both size and quality. Examples of imitative
FTR-FH coins are as small as 2-2.5 mm in diameter and 0.02 to 0.05 g (20 to 50 mg). One example as
small as 1.5 mm in diameter is known from excavations at Canterbury, UK, where it was identified
only by the fact that it was accompanied by another 154 FTR-FH copies that ranged from over 14
mm down to 2.5 mm.66 Several hoards from Britain are comprised entirely of FTR-FH copies with
average weights of 0.6 to 0.75 g.67
63
Brickstock (1987), 113.
64
Depeyrot (1992), 100.
65
66
67
Depeyrot (2001), 115.
144
Though these imitative coins could be used individually as small change, the smallest coins
were likely used by weight (being made up into small bags of a set weight) as they were too tiny to
use easily on their own. In this case, it was the weight of the entire bag, and not of the individual
coins, that mattered.68 Coins of different sizes could, therefore, be mixed together in a bag.
The imitations continued to be used long after they were created. Indeed, very small FTR-
FH imitations survived in hoards through the fifth century, when their small size was close to the
official standard of the contemporary small nummi.
The imitations were generally made of an inferior alloy compared to the originals. They
often have as much as 25% lead and 10% tin mixed with the copper. Many even have some silver,
though it is doubtful that this silver was intentionally added. Instead it likely came from the mixed
bronze alloy obtained from old melted bronze coinage.69
CONCLUSION
The FEL TEMP REPARATIO Falling horseman (FTR-FH) coin type is one of the most important Roman
coin types given its ubiquity in hoard and site finds throughout the Empire. A more refined
understanding of its history, and of the chronology of its six series, as provided in this article, is
important for a better understanding of Roman numismatics, and of late Roman history in general.
Hopefully, in the future, reports of hoards, site excavations and collections (both museum and
private) will benefit from being able to distinguish this coin type by series. Such data would assist
with the further study of the FTR-FH coin type itself – for example by providing weight, diameter
and fineness statistics broken down by series. In addition, reporting by series would provide
greater chronological refinement and would thereby assist us in gaining a better understanding of
this complicated and turbulent time period.
In fact, a recent article by Bagnall and Bransbourg may have provided the clue to unlocking
the meaning of the M on the sixth series of FTR-FH coins introduced likely circa mid-356. If indeed
this M represents the introduction of a new, and greatly increased, value of one myriad (10,000 d.c.)
per coin, then the numismatic evidence will have provided valuable insight into the inflation, and
overall economy, of the mid-fourth century.
It is also hoped that the breakdown of the design variations of the FTR-FH type as provided
in this article, particularly of the horseman design, will assist in facilitating the identification of
FTR-FH coins that are either worn or where the mint marks are no visible.
BIBLIOGRAPHY
Alram, Michael and Franziska Schmidt-Dick (eds). 2007. Numismata Carnuntina: Forschungen und
Material. 3 volumes. Vienna.
Bagnall, Roger S. and Gilles Bransbourg. 2019. ‘The Constantinian Monetary Revolution,’ Institute
for the Study of the Ancient World (ISAW) Papers 14. Accessed at http://dlib.nyu.edu/awdl/isaw/
isaw-papers/14/ on 2019-03-01.
Bastien, Pierre. 1985. Le monnayage de l’atelier de Lyon: de la mort de Constantin à la mort de Julien (337-363).
68
69
145
SHAWN CAZA
Beard, Mary, John North and Simon Price. 1998. Religions of Rome. 2 volumes.
Brickstock, R.J. 1987. Copies of the Fel Temp Reparatio Coinage in Britain. BAR 176.
Bruck, Guido. 1961. Die Spätrömische Kupferprägung: Ein Bestimmungsbuch Für Schlecht Erhaltene
Münzengung.
Callu, Jean-Pierre. 1986. ‘Aspects du quadrimestre monétaire,’ Mélanges de l’Ecole français de Rome.
Antiquité 98: 1.
Carson, R.A.G., Hill P.V. and J.P.C. Kent. 1978. Late Roman Bronze Coinage. (First published in 1960).
Caza, Shawn. 2018. ‘Redating Nepotian’s Usurpation and the Coinage of Magnentius’, KOINON I.
Cope, Lawrence H. 1974. The Metallurgical Development of the Roman Imperial Coinage during the First
Five Centuries A.D. Unpublished PhD Thesis, Department of Chemistry, Liverpool Polytechnic,
March 1974.
Crawford, Peter. 2016. Constantius II: Usurpers, Eunuchs and the Antichrist.
Depeyrot, Georges. 1992. ‘Le système monétaire de Dioclétien à la fin de l’empire romain,’ Revue
Belge de Numismatique et de Sigillographie CXXXVIII.
Depeyrot, Georges. 1996. Les emissions monetaires d’Arles (IVe-Ve siècles). Moneta 6.
Depeyrot, Georges. 2001. Les numéraire gaulois du IVe siècle: I Aspects quantitatifs, et II Les trouvailles,
2nd ed. Moneta 24 and 25.
Failmezger, Victor. 2002. Roman Bronze Coins: From Paganism to Christianity, 294 - 364 AD.
Ferrando, Philippe. 2010. L’Atelier Monetaire D’Arles: de Constantin a Romulus, 313 - 476.
Jones, John Melville. 1990. A Dictionary of Ancient Roman Coins.
Kent, J.P.C. 1981. The Roman Imperial Coinage, volume VIII, The Family of Constantine I, 337-364.
Kraft, Konrad. 1958. ‘Die Taten der Kaiser Constans und Constantius II,’ JNG 9.
Mattingly, Harold. 1977. ‘Fel. Temp. Reparatio,’ 1977 offprint from 1933 Numismatic Chronicle.
Metcalf, William. 2012. Oxford Handbook of Greek and Roman Coinage.
Sagramora, Moreno. 2010. Moneta Aqvileiensis (294 - 452).
Seeck, Otto. 1890. Die Münzpolitik Diocletians und seiner Nachfolger, Zeitschrift für Numismatik 17.
Stevenson, Seth William. 1889. A Dictionary of Roman Coins.
Vasić, Miloje R. 1978. ‘Le trésor de Boljetin (IVe siècle),’ Sirmium VIII. École Français de Rome.
Voetter, Otto. 1921. Die Münzen der römischen Kaiser, Kaiserinnen und Caesaren von Diocletianus (284)
bis Romulus (476): Katalog der hinterlessenen Sammlung und Aufzeichnungen des Herrn Paul Gerin
compiled by Otto Voetter.
146

FEL TEMP REPARATIO Falling Horseman Coinage Article

Uploaded by

Copyright:

Available Formats

FEL TEMP REPARATIO Falling Horseman Coinage Article

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FEL TEMP REPARATIO Falling Horseman Coinage Article

Uploaded by

Copyright:

Available Formats

KOINON

The International Journal of Classical

Alberto Campana, IT Mark Fox, US David Sear, US

Nicholas J. Molinari, General Editor

Published by Archaeopress Publishing Ltd, Oxford, UK

Subscriptions to KOINON should be sent to

Φαῖδρος: καὶ ἐμοὶ ταῦτα συνεύχου: κοινὰ γὰρ τὰ τῶν φίλων.

-Plato, Phaedrus (closing prayer), 279b-c

A Bacchid at Apollonia: a late survival of an ancient family�����������������������������������������������������������20

The Macedonian Mint At Susa (319/8-312/1 BC)����������������������������������������������������������������������������28

A discussion on provenance research with some early provenances uncovered��������������������������83

The Youthful God revisited: Veiovis on Roman Republican denarii�����������������������������������������������97

MEDIEVAL AND EARLY MODERN COINAGE

Numismatic letterforms of ‘A’ in medieval Europe: A classification system��������������������������������176

KOINON II 2019: 113–146

Failmezger’s 1992 Series Failmezger’s 2002 Series

Denomination Weight Standard Bust Faces Constans’ Type Constantius’ Type

Large 1/60 L Right Large Galley Falling Horseman

THE SIX FEL TEMP REPARATIO FALLING HORSEMAN SERIES

The first series: April 348 to mid-349

The special globe bust

The second series: mid-349 to the end of 350

Example of FTR-FH from the third series: Second issue

The FTR-FH Type under Magnentius: 350 to 351

MINT STRUCK FOR STRUCK FOR CONSTANTIUS II BY MAGNENTIUS

The FTR-FH Type under Vetranio: 1 March to 25 December 350

The third series: beginning of 351 to September 352

Examples of FTR-FH from the

The Fourth series: September 352 to November 353

S|- A on Obv, A on Obv, Δ on ●S●|✶

The Fifth series: November 353 to 356

E|- -|- II|- -|- -|- X●|-

The Sixth series: mid-356 to November 358

THE END OF THE FALLING HORSEMAN TYPE

The six FEL TEMP REPARATIO Falling Horseman Series Summary

Series Weight Average Marks Issue Struck For

THE DESIGN OF THE FEL TEMP REPARATIO FALLING HORSEMAN TYPE

Nineteenth-century Engraving of a Mithraic Tauroctony relief.

FEL TEMP REPARATIO FALLING HORSEMAN SUB-TYPES AND VARIATIONS

Sub-types based on the position of the horseman

FH1 Horseman in kneeling position beside horse

Sub-Type Horseman Design LRBC RIC-VIII

Sub-type A) Horseman facing Emperor

On this sub-type the horseman’s body faces the Emperor with

Sub-type B) Horseman leaning forward

On this sub-type, the horseman’s body leans forward, away from

Sub-type C) Horseman reaching back

On this sub-type the horseman’s body faces forward, but he turns

VARIATIONS BASED ON THE DESIGN OF THE HORSEMAN

a. Horseman is bare-headed and beardless

Horseman Type Horseman Likely Portrays Kent

Series 1. 1/60 L 2. 1/60 L 3. 1/60 L 4. 1/72 L 5. 1/120 L 6. 1/144

Ambianum -- -- -- Bi, Ciii -- --

VARIATIONS BASED ON THE POSITION OF THE EMPEROR

IMITATIONS OF THE FEL TEMP REPARATIO FALLING HORSEMAN TYPE

You might also like

A Bacchid at Apollonia: a late survival of an ancient family��20

The Macedonian Mint At Susa (319/8-312/1 BC)��28

A discussion on provenance research with some early provenances uncovered��83

The Youthful God revisited: Veiovis on Roman Republican denarii��97

Numismatic letterforms of ‘A’ in medieval Europe: A classification system��176