Thracian Units in the Roman Army
Author(s): M. G. JARRETT
Source: Israel Exploration Journal, Vol. 19, No. 4 (1969), pp. 215-224
Published by: Israel Exploration Society
Stable URL: https://www.jstor.org/stable/27925201
Accessed: 18-05-2020 17:05 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Israel Exploration Society is collaborating with JSTOR to digitize, preserve and extend access
to Israel Exploration Journal
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
Thracian Units in the Roman Army*
M. G. JARRETT
University College, Cardiff
The Roman army possessed many units raised in Thrace, both before and
after the creation of the Roman province in a. d. 46. They include both
cavalry (alae) and infantry (cohortes), in units five hundred or (in the case of
some cohorts) a thousand strong. Most, if not all, of these units appear to
have been f ormed during the first century a. d.; there may have been subse
quent Thracian numeri, but no attempt has been made to find evidence for
them. Unfortunately for the modern scholar, the Roman army was unmethodi
cal in its numbering and naming of units, and every time new Thracian units
were raised, they were numbered from cohors (or ala) I upwards; only rarely
was a distinguishing title added?and this might not be used consistently and
invariably. In consequence, there are at least five units which might appear in
inscriptions as ala I Thracum, and at least eleven with the title cohors I
Thracum. Given this fact, and the inadequate documentation which applies to
almost all units in the Roman army, it is scarcely surprising if confusion oc
casionally arises.
* The following abbreviations are used in this paper:
AE = L'Ann?e ?pi graphique.
BGU ? ?gyptische Urkunden a. d. K?n i gl. Museen zu Berlin: Griechische Urkunden.
Cheeseman (1914) = G. L. Cheeseman: The auxilia of the Roman Imperial Army, Oxford, 1914.
CIL = Corpus lnscriptionum Latinarum.
EE = Ep h emeri s Epigraphica.
ILS = H. Dessau: Inscriptiones Latinae Selectae, Berlin, 1892-1916.
Nagy (1956) = T. Nagy: The military diploma of Albertfalva, Acta Archaeologica Academiae
Scientiarum Hungaricae.
. D. = Notifia Dignitatum (ed. O. Seeck), Berlin, 1876.
RIB = R.G.Collingwood & R.P.Wright: The Roman Inscriptions of Britain I, 1965.
Richmond (1936) = I.A.Richmond: Roman leaden sealings from Brough-under-Stainmore, Trans.
Cumberland and Westmorland Antiquarian and Archaeological Society, NS
36 (1936), pp. 104-125.
E.Stein (1932) = Die kaiserlichen Beamten und Truppenk?rper im r?mischen Deutschland
unter dem Prinzipat, Vienna, 1932.
Wagner (1938) = Die Dislokation der r?mischen Auxiliarformationen in den Provinzen Nori
cum, Pannoni?n, Moesien und Dakien..., Berlin, 1938.
215
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
216 M . G. JARRETT
It was an example of this confusion1 which called the attention of the present
author to the problem. It cannot be suggested that problems have been or can be
solved: The most which the author may hope is that he has set forth the avail
able evidence in a comprehensible form, and has suggested some solutions, even
if most of the problems still remain. Two basic principles should be noted by
the reader: The simpler of alternative solutions has usually been preferred, and
it has been assumed that no unit is likely to have made frequent moves from
one end of the Empire to the other. This will normally have the result of sug*
gesting two units, where earlier writers?notably Wagner and Ernst Stein?
were prepared to believe in one. Occasionally the result has been the reverse of
this, and I have suggested that two units about which virtually nothing is known
may have been one and same. Future discoveries may well prove these sugges
tions wrong.
The principal sources for this study have been dated inscriptions and the
military diplomata (conveniently collected in Corpus Inscriptionum Latinarum
XVI), each listing units within a single province which had men due for
discharge on a given date, on completion of their twenty-five years' service.
Supplementary evidence comes from papyri, and from the early fifth century
Notifia Dignitatum. I have made no attempt to incorporate references to
inscriptions which have no bearing on the location of a unit, or the date at
which it existed. I have concerned myself entirely with the differentiation of
the various cohorts and alae, and their movements from province to province.
Earlier work on this subject begins with two monumental articles by Cicho
rius, ala and cohors, for Pauly-Wissowa: Realencyclop?die der das si sc hen
Altertumswissenschaft (RE). Published respectively in 1894 and 1901 these are
in need of revision, though still of great value. More recent regional studies
of the Roman army to which reference should be made are those of Wagner
(1938) and E. Stein (1932). I am indebted to Mr. R. P. Harper and Mr. R. P.
Wright for answering questions which have arisen during this study, and to
Dr. J. C. Mann for suggesting a number of improvements. This paper should
be read in conjunction with his note printed above.
No attempt is made to discuss units whose title includes not only Thracians
but another ethnic group. Nor does it seen necessary to follow the fortunes
of the ala I Thracum Herculiana, attested in Syria in 157.2 It seems clear that
1 IE] 17 (1966), pp. 52-53. 2 CIL XVI, 106.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
THRACIAN UNITS IN THE ROMAN ARMY 217
the unit was frequently known as the ala (I) Herculiana, never as the ala
Thracum without the distinguishing epithet Herculiana3 One series of a
with the additional title Augusta attached to each, can be traced.
1) ala I Augusta Thracum. Recorded at Gerasa (Syria), probably late in th
first century,4 this unit subsequently served on the Danube. It is atteste
only one diploma, that for Raetia of 107.5 It is missing from subsequen
Raetian diplomata. By 140-144 it was in Noricum,6 and may have been the
a little later.7
2) ala II Augusta Thracum. Attested in Mauretania Caesariensis in 1078 and
on many undated inscriptions from that province. Nothing further is known
of the unit.
3) ala III Augusta Thracum c. R. Attested in Syria in 88 and 91,9 this unit was
in Pann?nia Superior by 133.10 It is probably the ala III Thracum in Syr\id\ of
ILS 2711 = CIL II 4251. The Caracallan change in the boundary between the
two Pannonias left this unit in Inferior, where it remained until at least 252.11
4) ala I Thracum Victrix. Attested in Pann?nia Superior from 133-154,12 and
possibly as late as 163.13 The unit is not otherwise known, though ILS 2210
suggests that it may have been in Pann?nia as early as the reign of Trajan.
5) ala I Thracum Mauretana. Service in Mauretania (probably during the
conquest under Gaius and Claudius) may safely be presumed from the title
Mauretana. The title was certainly won before 86, when we find the unit in
Iudaea.14 Subsequently it is attested in Egypt from 142 or earlier to at least the
reign of Severus.15
3 E. g. ILS 2724, attesting its contribution to the force of M. Valerius Lollianus in Mesopotamia,
probably under Verus.
4 AE 1930, pp. 89-90; cf. C. B. Welles in C. H. Kraeling (ed.) : Gerasa City of the Decapolis, New
Haven, 1938, pp. 446-47. 5 CIL XVI, 55. 6 CIL III, 5654.
7 CIL III, 5655; cf. E.Stein (1932), pp. 154-5. 8 CIL XVI, 106.
9 CIL XVI, 35; AE 1961, 319. 10 CIL XVI, 76; 77; 96; 97; 99; 104; 178; AE I960, 21.
11 ILS 523 = CIL III, 4270. 12 C7L XVI, 76, 77, 84, 97, 104, 178.
13 AE I960, 21; the title Victrix is omitted. 14 CIL XVI, 33.
15 AE 1948, 56 (142); 447; EE 7, . 457 (154-6); ILS 2543 (199); 715 4424, dated to ca.
203-210 by the career of Subatianus Aquila, cf. A. Stein: Die Pr?fekten von ?gypten . . ., Berne, 1950.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
218 M. G. JARRETT
6) ala 1 Thracum vet er. sagittariorum. Nagy (1956) suggests, on the basis of
the names of men serving with this unit, that it may well have been in Pan
n?nia Superior before its transfer to Pann?nia Inferior. It is attested in the
latter province by diplomata for the period 139-167, and was still there under
Caracalla (or Elagabalus) and Severus Alexander.16 The latest record is a
milestone of 251.17 In 150 it had just returned from service in Mauretania.18
Nagy is probably correct in reading veter(anorum) rather than et er (ana),
since the latter would imply two units of the same name and number in the
province, and only one is attested.
7) alai Thracum. Attested in Britain in 103 and 124,19 it was probably in the
province earlier: an inscription from Cirencester, probably of the mid to late
first century,20 records an ala Thraec (sic), and no other ala Thracum is attested
in Britain. It is probably the ala I Thracum known in Lower Germany, but not
certainly there before the mid-second century;21 it was still in Germany in 219?22
Cichorius sought to identify this with one of the alae I Thracum serving on the
Danube in the second century;23 he is followed by the editors of RIB. The stud
ies of the other units (Nos. 1, 4 and 6, above) make this extremely unlikely.
8) ala II Thracum. Attested in Pann?nia about the year 100, this is probably
the unit of which a trooper was buried at his home town in Lusitania.24 Noth
ing further is known.
9) ala III Thracum. The existence of this unit is not certain. The ala III Thra
cum in Syria25 is surely the ala III Augusta Thracum. M. Macrinius Avitus Cato
nius Vindex served as prefect of an ala III Thracum in the later second century;
once again the ala III Augusta Thracum may be in question.26
COHORTES
Two series of cohorts can be readily distinguished, bearing respectively th
epithets Augusta and Syriaca. Other units present a more confusing picture.
16 cil III, 3394; 3388. 17 cil III, 10624. " cil XVI, 99.
19 cil XVI, 48; 70; cf. Nagy (1956). 20 rjb 109.
21 CIL XIII, 8818, 12058. This unit does not seem to merit a separate discussion.
22 cil VI, 31162, cf. E. Stein (1932), pp. 153-4. 23 R? 1 (1894), pp. 1263-4.
24 cil II, 812. Unit 2 may be in question.
25 7L5 2711 = cil II, 4251. 26 ils 1107 = C/L VI, 1449.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
THRACIAN UNITS IN THE ROMAN ARMY 219
10) Cohors I Augusta Thracum.c.R.eq. The unit is attested on four undated
inscriptions from Arabia, two from Motha, and one each from Kurnub and
Umm el-Qottein; that from Kurnub is likely to be of the early second century.27
The unit is attested in Pann?nia Inferior in 167,28 and may well have been
transferred to the Danube front after the Parthian War of Marcus and Verus.
Like the third cohort of the series (and possibly the second) it may well have
been raised for service in the East.
11) cohors II Augusta Thracum eq. Recorded in Pann?nia Inferior in 139,
148, 151/60 and 167.29 It was still in existence in 212-20.30
12) cohors III Augusta Thracum. Like the earlier cohorts of the series, this
unit was probably equitata (part-mounted). It is attested on Syrian diplomata
for 88 and 157.31
13) cohors I Thracum Syriaca eq. Presumably this, with the other units in the
series, was raised for service in Syria. Wagner suggests that it may have been
in that province under Claudius,32 By 78 it was in Moesia, and the division
of the province left it in Moesia Superior, where it is attested in 100.33 The
last record of the unit is the career of a prefect of ca. 200.34 The cohort could
be the same as cohors I Thracum sagittariorum, in Dacia Superior from 120 ( ?)
to 158.
14) cohors II Thracum Syriaca. The unit is attested on diplomata for Syria of
91 and 157,35 but otherwise unknown.
15) cohors III Thracum Syriaca. Recorded on the Syrian diploma for 88.36 The
career recorded on AE 1911, 161 suggests that the unit may have remained in
the East.
27 CIL III, 109, HO; IE] 17 (1967), p. 52; AE 1928, 254; cf. Dr. Mann's note above. All the
inscriptions from Arabia will, ipso facto, be later than 106.
28 CIL XVI, 12. For the titles civium Romanorum equitata, see ILS 9471.
29 AE 1955, 17; CIL XVI, 179, 180, 112, 113, 123.
so ILS 9014, cf. Nagy (1956) 68. 31 CIL XVI, 35 and 106.
32 I can find no positive evidence for this suggestion, though it is not unlikely.
33 CIL XVI, 22, 46. 34 AE 1926, i50.
ss AE 1961, 319; CIL XVI, 106. 36 CIL XVI, 35.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
220 M . G. J ARRETT
16) cohors IV Thracum Syriaca eq. While probably raised for service in Syria,
the unit was in Spain during the first century.37 Wagner (1938) produces
evidence, which is not conclusive, to suggest subsequent service in Moesia.
Cheeseman (1914) was not justified in supposing the unit to be in Syria in
117-161. Its survival into the second century is indicated by CIL II, 1970.
17) cohors I Thracum eq. This cohort is recorded on two (and possibly
three) diplomata for Pann?nia Inferior, which cannot be precisely dated.38 It
cannot be certainly equated with any of the other units in Pann?nia Inferior;
it is just possible that it is / Augusta Thracum (No. 10).
18) cohors I Thracum sagittaria. Attested in Dacia Superior in 144-158,39 but
probably not in 120,40 nothing further is known of this unit. It might be equat
ed with any one of two or three units whose movements in this period are not
known, but there is no positive evidence.
19) cohors I Thracum eq. Recorded in Germany on a tombstone from Cologne
which should probably be assigned to the first century41 and on the diploma
for Germania Inferior of 80,42 this unit had moved to Britain by 122,43 and
was still there under Severus.44 It is clearly not the same as the unit attested in
Egypt in 127 (No. 22). We may note a cohors II Thracum moved from Ger
many to Britain between the mid-first century and 103 (No. 28).
20) cohors I Thracum. A unit with this title is recorded in Iudaea in 86.45 It
may be the same as the cohors I Thracum attested at Qasr el-Hallabat (Arabia)
in 212,46 and at Asabaia, Arabia, when the Notifia Oignit at urn was compiled
eaily in the fifth century.47 It must be carefully distinguished from No. 21.
37 AE 1963, 27.
38 CIL XVI, 112, 113, both of 151-160; possibly also CIL XVI, 131, of 139-f*. 190.
ss CIL XVI, 90, 107, 108.
40 CIL XVI, 68; cf. AE 1958, 30, and J. C. Mann in Hermes 82 (1954), pp. 501-506.
41 ILS 2569. 42 C/L XVI, 158.
43 CIL XVI, 69; d.RIB 1323. But #LB 291 (Wroxeter) is possibly pre-Flavian; the unit attested
in Germany may, therefore, be better equated with No. 17.
44 RIB 730; 740. 45 CIL XVI, 33.
46 Cited in IE] 17 (1967), p. 52; for the reference see above, n. 4.
47 N.D. Or. 37, 32; Dr. Mann's suggestions (above, p. 211), which have much t
eliminate this unit.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
THRACIAN UNITS IN THE ROMAN ARMY 221
21) c oh or s I Thracum milliaria. First attested on the Syrian diploma for 91,48
this is presumably the unit attested at En-Gedi (Iudaea) in 124.49 In 139 it
was in Iudaea,50 and by the early fifth century was in garrison at Adittha,
Arabia.51 It is possible, but unlikely, that it is the cohort which was in Iudaea
in 86; the title milliaria is not recorded for that unit, and seems to be invariably
expressed for this one?probably to distinguish the one from the other.
22) cohors I Thracum. Attested on an undated inscription from Ara?
(Galatia) ,52 this cohort is otherwise unknown, unless it is unit No. 20 or, more
probably, the (otherwise unknown) cohors I Thracum equitata which was in
Egypt in 127.53 An inscription from Upper Moesia54 refers to a duplicarius of
cohors I Thracum equitata of Galatian origin: this might suggest that No. 22
is in question.
23 & 24) cohortes I Thracum c.R. It is clear that there are two units (at least)
of this title, whose movements cannot now be disentangled. One was in Moesia
Superior in 100 and Dacia in 110.55 It was moved either to Pann?nia Superior
by 133,56 remaining until 154 or even 163,57 or to Pann?nia Inferior by 148;58
this latter is probably the I Thracum c.R. which was under the dux Pann?ni??
Secundae early in the fifth century.59 This unit may have been in Pann?nia
Inferior as early as 110, when we find a cohors I Thracum c.R. p.f. which is not
otherwise attested. It may appear on other diplomata for the province as cohors
I Thracum [ ], in the period 151-160.60 Over these two units and
No. 25, my suggestions differ radically from those of Wagner (1938) and
E. Stein (1932).
25) cohors I Thracum Germanica c.R. The honour Germanica was presum
ably won in one of the German provinces, and we may therefore reasonably
equate this unit with the cohors I Thracum c.R. recorded on diplomata for
Germany (and subsequently Germania Superior) from 74 to 116; the title
48 AE 1961, 319; probably the coh. I mill/aria in Syria in 88 (CIL XVI, 35, cf. Dr. Mann's note,
above).
49 IE] 12 (1962), p. 259. 50 CIL XVI, 87. 51 . D. Or. 37, 31.
52 AE 1951, 254. 53 AE 1937, 112. 54 ^? 96^ 12.
55 CIL XVI, 46, 163. 56 CIL XVI, 76. 57 CIL XVI, 104; for 163, see AE I960, 2
58 CIL XVI, 179, 180. 59 . D. Or. 32, 59. 60 C/L XVI, 112, 113.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
222 M. G. JARRETT
civium Romanorum is recorded on only the document for 116.61 As I Thracum
Germanica c.R. it is recorded in Pann?nia Inferior in 148 to 167,62 and again
in 235-38.63 Wagner (1938) equates it with the cohors I Thracum c.R. attested
in Moesia* Superior in 100 and in Dacia in 110 (No. 23); neither of these
units has the title Germanica, and it seems better to treat them separately.
Wagner's attempt to reduce the number of 1st Cohorts carries simplification
to the point of confusion; it involves the unit serving in six provinces (Ger
many, Upper Moesia, Dacia, Upper Germany, Upper Pann?nia, Lower Pan
n?nia) between 90 and 151. The move to Pann?nia Superior must certainly be
omitted from this sequence: the unit in question (No. 23-24) was in Pann?nia
Superior from 133 to 154, while / Thracum Germanica is now known to have
been in Pann?nia Inferior by 148.64 / Thracum Germanica seems likely to have
had a less adventurous career than has previously been supposed, serving in
Germany and Upper Germany before moving to Lower Pann?nia at some date
between 116 and 148.
26) cohors II Thracum c.R. Cohors II Thracum is attested in Iudaea in 86,65
and may be the same as the cohors II Thracum c.R. which was in Syria in 88
and 91.66 A unit of the same name, probably the same cohort, was in Egypt
in 167 and in the early fifth century.67
27) cohors II Gemella Thracum eq. This unit is attested in Numidia, though
the inscriptions cannot be precisely dated.68 There seems to be no other record
of it.
28) cohors II Thracum. A career inscription from Aquileia reveals that this
cohort was in Germany at a date which might be earlier than the accession of
Claudius. By 103 it was in Britain, where it is attested in 122, under Pius, and
at whatever later date the Notifia sub-section per lineam Valli was compiled.69
Perhaps the most likely dates for its transfer to Britain are 61 (after the
Boudican revolt) or 71 (with leg. 11 Adiutrix), but it may have been part of
the invasion force of 43.
ei CIL XVI, 20, 28, 36, 62, 63 (?). 62 CIL XVI, 179, 180, 112, 113, 123.
63 CIL III, 3750. 64 CIL XVI, 179, 180. es c/L , 33.
66 CIL XVI, 35; 1961, 319- 67 Wilcken: Ostraka 927; N.D. Or.
68 E. g. /LS 2578 = CIL VIII, 2251, 5885 + p. 964.
69 ILS 9090; CIL XVI, 48, 69; RIB 2142; . D. Of. 40, 50.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
THRACIAN UNITS IN THE ROMAN ARMY 223
29) cohors III Thracum veterana.
30) cohors III Thracum c.R. eq. bis torquata. One or other of these units
presumably that recorded in Pann?nia in 80-84; probably No. 29, since t
Pannonian records do not include the title equitata. Both are atteste
diplomata for Raetia for the period 107-166.71
31 & 32) cohortes IV Thracum. Two cohorts are implied by the existenc
two sixth cohorts. One was in (Lower) Germany in 80,72 and one apparen
in Lusitania at an uncertain date.73 One or other is attested on three car
inscriptions of the pre-Flavian period.74
33 & 34) cohorts V Thracum. Neither of the two fifth cohorts is epigraphical
attested, and their existence (perhaps brief) has to be deduced from the
that two sixth cohorts are almost certainly known.
35) cohors VI Thracum. Attested at Gloucester (Britannia), probably in
middle of the first century, the unit was in Britannia Inferior in the th
century.75
36) cohors VI Thracum. eq. Recorded in Lower Germany in the Claudian
period and in 80,76 this cohort subsequently appears in various Danubian
provinces from 84 to 164: Pann?nia in 84 and 85 ;77 Upper Moesia in 100 ;78
Dacia in 110;79 and Dacia Porolissensis in 159 and 164.80 It is just possible,
as Wagner thought, that Nos. 35 and 36 are one unit; but it seems unlikely.
37) cohors VII Thracum. Only one seventh cohort is known. It was in Britain
in 122 and 135,81 and in Britannia Inferior in the third century.82
From this study, we may perhaps reconstruct another series of Thracian
CIL XVI, 26, 50. 71 CIL XVI, 55, 94, 117, 118, 121, 125, 183, 187; AE 19?, 105.
72 CIL XVI, 158, cf. CiL XIII, 7050, 7585.
73 CIL II, 4212; cf. A.Garcia y Bellido in Archivo Espa?ol de Arqueolog?a 34 (1961).
74 ILS 2706 = CIL XIV, 3548; CIL II, 4138, 4212.
RIB 121; Richmond (1936). 76 CIL XIII, 7052; CIL XVI, 158.
77 CIL XVI, 30, 31. 78 CIL XVI, 46. 79 CIL XVI, 163.
*o CIL XVI, 110, 185. 81 CIL XVI, 69, 182. 82 Richmond (1936).
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
224 M. G. JARRETT
cohorts, raised for service in Germany, and numbered from I to VI or VII. It
consists of:
I: Unit 19, in Germany until 80 at least, after which date it moved to
Britain or Pann?nia Inferior (Unit 17).
II: Unit 28, in Germany in the mid-first century, but in Britain by 103.
Ill: Unit 29 or 30, neither of which is attested in the first century, though
at least one must have been in existence.
IV: Unit 31, in Lower Germany in 80, but thereafter unknown.
V: Unit 33 or 34.
VI: Unit 36, in Germany from the Claudian period to just later than 80, and
thereafter on the Danube.
VII: Unit 37, first attested in Britain in 122. Like other Thracian cohorts in
Britain, previous service in Germany is not unlikely.
Of these cohorts, I and VI were certainly equitatae, and III may well have
been; it is by no means improbable that the whole series was part-mounted.
At least ten, and probably thirteen (if we include the whole of the Augusta
and Syriaca series) of the twenty-eight certainly known cohorts were equitatae,
a higher proportion than we should expect amongst the auxilia of the Roman
army. In addition we may note two units of archers, one cavalry and one
infantry; once again the proportion is higher than we should expect, and has a
bearing on the fighting methods of the Thracians in the early first century,
when most of these units were probably raised.
This content downloaded from 193.19.172.29 on Mon, 18 May 2020 17:05:29 UTC
All use subject to https://about.jstor.org/terms
