1 Ancyra, Metropolis Provinciae Galatiae

Julian Bennett

/11troductio11 Tamerlane- and Kaledag1 (Figure 1.2). Exactly when this

location was first settled is uncertain. The Tamerlane- and
John Wacher's achievements in elucidating the Roman Kaledag1 do, however, constitute a conspicuous way-point
heritage of many of modern Britain's urban centres has on the northern trans-Anatolian highway, a trade-route in
hardly been equaled for the other regions of the Roman use from at least the early 2nd millennium BC, and we
Empire. This is especially true of Asia Minor, where
might assume that some settled occupation existed here
secure information about the more than I00 urban from an early period. Such is indeed suggested by the
settlements of classical date now sealed beneath the place-name itself, for the locative 'Anku-' is well-attested
conurbations of today's Turkey is woefully inadequate, regionally in Hittite times (e.g. Ankuva, Ankuwash and
for only in the mid- I990s did the value of urban archae­
Ankuruva). It emanates from *ango, a word of autoch­
ology began to be properly recognised. Nowhere is this thonous Indo-European origin meaning an elbow-shape,
deficiency more obvious than in the case of Ancyra, from whence the Greek and Latin equivalents ankon/
metropolis provinciae Galatiae, and precursor of the ancon. From these two explicit labels evolved: the sailor's
Turkish Republic's capital of Ankara. Despite some 80 ankora/ancvra, or anchor; and the toponymic ankona/
years of near-continuous urban development, our know­
ancona, a landscape feature of crooked or elbow form,
ledge of its early history relies essentially on anecdotal such as the V-shaped bay which gives its nan1e to Ancona
evidence, for few controlled excavations have been in peninsular Italy.
carried out here, and none have yet been published in a As it was, the citizens of Ancyra explained the origins
satisfactory way. That said, as John Wacher has shown,
of their po/is with reference to the first of these deriva­
even the most ephemeral information does allow some tives, pointing to an anchor displayed in their Temple to
commentary, and his efforts in this regard with the towns
Zeus and frequently depicted on the local coinage (Paus.
of Roman Britain serves as the direct inspiration for this 1.4.5; Arslan 1991, pls. 1.1, I O and 1.9, C I). One account
paper.
claimed this had been found by Midas, the semi-le.gendary
ruler of Phrygia in the late 8th century BC, and prompted
the name for the settlement he subsequently founded at
Pre-Roman A11cyra the find spot (Paus. 1.4.5; cf. Tzetzes Chi/. 1.139, and
Lib. Ep. 1230; Mitchell 1993a, 19-20). According to
The Central Anatolian Steppe (Figure 1.1) is a vast arid another, the anchor was a war trophy presented by the
expanse of acidic and saline soils, exposed to bitterly cold kings of Pontus and Cappadocia to the Galatian Tect­
winters and scorching summers, and in pre-industrial osaoes for their cooperation against a Ptolemaic attack on
times suitable only for rough grazing by sheep and goats. ::,
.
the Pontus: on return to Galatia they founded their first
In places, however, the rolling landscape is interrupted by po/is, naming it for their prize (Mem. FGrH 536; Steph.
prominent basalt outcrops, which offer favourable lo­
Byz. FGrH 740. F 14; Mitchell 1993a, 19-20). Both of
cations for permanent occupation. They provide ready these fables, however, were evidently conceived in
access to water from springs and snow-fed rivers and
Hellenistic times, when the region's indigenous poleis
streams, and also some limited protection from the wind contrived ·foundation myths' to assert a long-standing
and sun - and the attacks of any enemies. affinity with the Hellenised world and obscure their
Modern Ankara occupies just such a location, where barbarian origins (Jones 1940, 49-50). · Ancyra' is
the Ankara <;ay exits a 120111 deep gorge between the assuredly a toponymic, describing the course of the
2 Julian Bennett

ANCYRA
•

Q,,�
if;
��

Figure I. I Regional location map for Figure 1.2.

Ankara <;ay as it makes an almost 90 degree change in Anab. 2.4.1). While neither source is contemporary with
direction on leaving the gorge between the Tamerlane­ the event, both writers used earlier material which was.
and Kaledag1 (Figure 1.2). We might therefore assume some form of permanent
Although the toponym might suggest Hittite occu­ settlement at Ancyra at the time, perhaps on the Kaledag1,
pation, the earliest secure evidence for activity at Ancyra a natural acropolis, and a region yet virgin territory in
belongs to the Phrygian period. Sherds of 8th-5th century archaeological terms. That apart, almost another century
Phrygian pottery have been found in the Ulusdag1 and was to pass before Ancyra again featured in the historical
<;ankmkap1 districts, along with the remains of con­ record, as the place where Seleucus II was defeated by his
temporary buildings (Ozgii<;: 1946, 557-97; Dolunay brother Antiochus Hierax in c. 240/239 BC (Pomp.Trog.
1941, 263; Metin and Akahn 1999). More significantly, Prof. 27). Then, some 50 years later, Ancyra makes
at least I O tumuli of Phrygian type exist(ed) within a I 0km another appearance in documentary sources, when M.
radius ofUlusdag1, indicating that a substantial settlement Vulso camped by the castris and 'notable' urbs of the
stood hereabouts in the mid-I st millennium (e.g., Ozgii<;: Galatian Tectosages (Livy 38.24.1-25.1, cf. Poly. 21.39.
and Akok 1955). If so, the lack of any later Iron Age 1-2).
material from Ancyra might be explained by the decline Archaeological evidence for any form of activity at
in importance of the trans-Anatolian highway after the Ancyra in the Hellenistic period is, however, virtually non
collapse of the Phrygian Empire in c. 675 BC, accelerated existent: a few coins of late 4th - early 2nd century date
during the Persian occupation of Anatolia, when most and some 'Hellenistic' pottery, all this material being
trade followed the Royal Road along the southern edge of found in (?)residual contexts in the Ulusdag1 districts
the Anatolian Steppe. (Arslan 1996, I 08; Krencker and Schede 1936, 46;
Be that as it may, the northern route was chosen by Temiszoy et al, 1996). Historical considerations, on the
Alexander the Great for his march across Anatolia in 333- other hand, indicate there was some form of permanent
332 BC, when Ancyra first enters the historical record - settlement here at least by the end of the I st century BC:
albeit retrospectively. According to Roman sources, the when provinciae Galatiae was formed in c. 25 BC, all
Macedonian army encamped here while Alexander took three Galatian tribes were assigned a central meeting
the surrender of Paphlagonia (Curt. Alex. 3.1.22; Arr. place, Ancyra for the Tectosages, Pessinus for the
 I Ancyra, Metropolis Provinciae Galatiae 3

NORTH GALATIA

• • Roman Cities & Towns

-
Province boundary

''· - . -.. . ,... .. . .. / Roman Road

'· /

·t�
Find Spots
Rivers

- -,"(} 50 KM
OrawJ' by B Claasz Coockson 2002

Figure 1.2 Galatia, ·with principal Roman settlements indicated, and distribution of epigraphic find-spots of Roman
date; based on Mitchell 1982a, endpaper illustration.

Tolistobogii, and Tavium for the Trocmi (Mitchell 1993a,

Early Roman Ancyra
87). At this time, Pessinus was a thriving temple-state, The inaugural governor of provinciae Galatiae was M.
while Tavium was both a sanctuary to Zeus and an Lollius (Remy 1989, 127-29). He was presumably
emporion (Strabo 12. 5.2-3). We might assume, therefore, responsible for the first stages in transforming Ancyra
that some form of permanent settlement also existed at into a fitting urban centre for the Galatian Tectosages: the
Ancyra - perhaps even the nea po/is of (?)Arsinoe allocation of a suitable territorium, and the introduction
founded by the Galatian ruler Deiotarus in c. 54/53 BC of an appropriate political constitution and infrastructure.
(Plut. Cras. 17.1-2; cf. Mai. Chron. 9.221 (ed. Dindorf), We can calculate the territorium of Ancyra as an area of
for the possible name). Unfortunately, classical sources some 25, 000 ha (Figure 1.2), which by the mid-2nd
are ambiguous on the nature of Ancyra at the time. Strabo century AC contained some nine lesser settlements (Ptol.
categorises it as a po/is in one place, and elsewhere Geog. 5.4). Find-spots of inscriptions within this area
describes it as a phourion, a fortress (4. 1. 13 (187) and suggests that there were some 50 or more smaller villages
12.5. 2 (567)). Pliny, using late I st century BC material, and farmsteads by the end of the 4th century AC, many of
simply designates Ancyra as an oppidum (NH 5. 46), his which are likely to have their origins in the pre-Roman
favoured portmantologism when Jacking detailed in­ Iron Age (e. g. , Yalmcak: Tezcan 1964; 1966; this paper,
formation about a given place. It must be conceded, Figure 1.2). Most of these rural communities are located
therefore, that in the light of current evidence, the in the southwest and northwest of Ancyra's territorium,
character of Ancyra before the annexation of Galatia where there are extensive areas of land suitable for the
remains moot. Consequently, for the purpose of this paper, grazing of sheep and goats. Indeed, throughout the pre­
we shall follow the orthodoxy in assuming that all remains industrial period, wool would appear to have always been
of the classical period at the place are no earlier than c. the main source of the region's wealth, Amyntas, Galatia's
25-22 BC (cf. Mitchell 1993a, 86-7). last king, for one owning 30 flocks, while Pliny and others
4 Julian Bennett

make reference to the high quality of Galatian cloth and it seems that at least during the reigns of the first four
the local natural dye-stuffs (Strabo 12.6.1 (568); Pliny Julio-Claudian principes, these men were rewarded with
NH 9.141, 16.32, 22.3, and 29.33; cf. Exp.tat.11111nd.gent. Roman citizenship (e.g. Bosch 1967, nos.55 and 98). That
41 (ed. Rouge); also Mitchell 1993a, 146). As early as the apart, we can assume that Ancyra also had the other usual
16th century, English wool merchants began to migrate to annual magistracies of a large paleis, the tamiai, an
what was then known as Angora to capitalise on this ekdikas, and a gymnasiarchan, although epigraphic
commodity (French 1972; cf. Barnett 1974), but'Angora evidence for these is lacking. By contrast, there is plentiful
wool' is now available world-wide as'mohair', and today, record of the irregular magisterial post of palitagraphas,
a mere five or so goats at Ankara's 'Model Farm of who in the eastern paleis under Roman jurisdiction
Atatiirk' testify to this once vital Galatian staple. registered those citizens eligible to benefit from the will
As for Roman Ancyra's political constitution, in­ of an imperial freedman (Bosch 1967, nos.249-253 and
scriptions reveal this to have been based on the Hellenic 287-288; cf. Mitchell 1977, no.7, and p.74). As such, the
model, with a demas formed from among its free-born repeated need for such a post at Ancyra confirms the
citizens (cf. Bosch 1967, no.72; Mitchell 1982a, no.178). regular deployment of imperial freedmen to the admin­
The people were divided on geographical grounds into 12 istration of the province (e.g. Bosch 1967, nos. 64-65 and
phylai for administrative purposes, and it has been 276).
suggested from their names that there were six initiatory Matching the political institution of the baule was the
phylai, the/ Maururagene, II Pakalen/e, Ill Menari::eitan, religious association of the imperial cult, headed by an
IV Hiermene, V Dias Trape::an and VI Sebaste; then two annually elected archiereus, entitled by his rank to wear
more were added under Claudius, the VII (?)-mene and purple garments (Mitchell 1977, 6). He appears to have
VIII Claudia Athenaea; another two under Nerva, the IX had an associate, the Ga/atarchan, whose precise function
Hiera Bulaia and the X Nerva; and then, probably under is unclear (Mitchell 1977, 7). However, it is clear that
Hadrian, a final two were formed, the XI Nea O(vmpias Ancyra was the main location for the koinan of the
and XII Dias Taenan (Mitchell 1977, 80-1). If this Galatians, an organisation probably established under
restored sequence does indeed reflect the actual situation, Tiberius, and which maintained the imperial cult and its
it indicates that Ancyra doubled its population between its associated festivals. The latter involved the archiereus in
foundation and the reign of Hadrian (cf. Mitchell 1977, substantial and lavish expenditure, sometimes on buildings
81). On the other hand, it may be that the original and statues, but mainly on more ephemeral benefactions,
constitution envisaged 12 phy!ai, a common number in such as public shows and banquets, and donations of olive
many Hellenic paleis (cf. Jones 1940, 158; Plato, oil and grain to the populace (cf. Bosch 1967, 51; Mitchell
Laws,737 and 745), and that some were later renamed for 1993a, I 08). By the time of Nero, the Galatian kainan
one or other reason. That apart, little can be said had replaced this method· of largesse with the mega/a
concerning the phylai other than each was headed by an Augusteia Actia, a four-yearly cycle of Hellenic-style
elected phylarchan, with an elected astynamas respon­ 'games', supervised by an elected or nominated agana­
sible for maintaining the streets and sewers in his ward thetes (Moretti 1953, no. 65; Bosch 1967, no. 287; cf.
( e.g., Bosch 1967, no. 20 I; Mitchell 1977, no. 9). Robert 1960). Other evidence indicates that at least two
The principal administrative organ of Roman Ancyra further aganes were introduced in later years, the first the
was its baule (cf. Bosch 1967, no. 72). The size and status aganes mystikai, an artistic festival dedicated to Hadrian,
of the place suggest there were at least the usual 500 and inaugurated at Ancyra in the emperor's presence on 7
bau/eutai, but while the bauleutai in Hellenic pa/eis were December 129 (Bosch 1967, nos. 127-130; cf. Oliver
elected by the phylai, in Ancyra they were chosen from a 1989, 96A-C). At a later date came the mega/a Isapythia
strictly defined social class, as normal in the Roman East Asclepieia Satereia (Antaneineia), probably established
for those paleis formalised after annexation. This is during the reign of Caracalla on the initiative of Titus
demonstrated by the existence of baulagraphai, the Flavius Gaianus, an Ancyran ambassador to that princeps
censors who listed those citizens who qualified for the (Bosch 1967, nos. 285-286; cf. Mitchell 1977, 7 and 8).
bau/e - evidently by property prerequisites, thus placing There are some indications that Roman Ancyra was a
the government of Ancyra firmly in the hands of its planned settlement with an orthogonal layout. Such may
wealthier citizens (Bosch 1967, nos.287 and 288; cf. Jones be deduced from von Yincke 's 1839 plan of Angora
1940, 171; Pliny Ep. 10.112-114). (Eyice 1971, pl. 39), for certain street and property
The chairmanship of the Ancyran bo11/e was vested in alignments appear to follow a regular north-south and
a single archan (cf. Bosch 1967, no. I00), presumably east-west pattern. It is a reasonable assumption that this
elected on an annual basis. The executive arm was street-plan originated in an earlier pre-medieval layout, an
likewise probably elected on an annual basis, and assumption somewhat reinforced by the apparent coinci­
inscriptions attest to three regular junior magistracies: the dence of one of these 19th century streets with a classical­
agoranomos (cf. Bosch 1967, no. I03); the bo11/agraphas period north-south street at Ulus Meydan1, and in the way
(Bosch 1967, no.289); the eirenarchan (Bosch 1967, that at least one building in Ancyra, the bath-house on the
no. I00). On the evidence of their personal nomenclature, Askeri Cezevi site, is aligned exactly north-south and
 I Aneyra, Metropolis Provinciae Galatiae 5

west-east (Temiszoy et al, 1996, figs. I and 2; Akok 1955, elaborate than a defined space with a plinth supporting
fig 9). From this admittedly circumstantial evidence, we statues (cf. the sebasteion at Bubon: inan 1994, I06 (but
might tentatively suggest that Ancyra was laid out with cf. Haley 2000, 28-9): note that the fifth arehiereus,
insulae in the order of 140-160111 square (cf. Figure 1.3). Albiorix, donated statues of (Tiberius) Caesar and Julia
If so, however, there is one clear exception to an entirely Augusta to the po/is, perhaps for the sebasteion).
orthogonal plan, a colonnaded street with a northwest - Although we do not know who was responsible for
southeast alignment discovered in I931 immediately building the 'Temple of Augustus', or when, the 'priest
northeast of the 'Caracallan Baths' (Dalman 1932, 122- list' and other epigraphic evidence indicates that it was
33; this paper, Figure 1.3). We might conjecture that it the arehiereis and the officials of the bou/e, along with
was based on a pre-existing route, presumably the ancient the agonothetes, who played the leading part in the
trans-Anatolian highway on its way from Ancyra towards Romanization and thus the urbanisation of Ancyra. The
Gordion, via a ford or bridge over the Ankara <;ay. 'priest-list', for example, records how several arehiereis
Von Yincke's plan also indicates that the focus of made substantial gifts of olive oil to the po/is, the first in
Ottoman Angora was the open space now represented by e. AD 20/21, suggesting that a gymnasium - a defining
Hi.iki.imet Meydan1 (Figure 1.3). This space may well be feature of Hellenised life - existed by then (Bosch 1967,
the direct descendant of Ancyra's agora, for not only was no. 51, lines 7-8; cf. Mitchell I993a, I08). Moreover,
it dominated in classical times by the so-called Augustus from the same date, there are several references in the
Mabedi, or 'Temple of Augustus', but the reputed'Column 'priest list' to the entire range of speetaeulae - gladiatorial
of Julian' (in reality probably a 6th century monument: cf. (one event involved 50 pairs), equestrian (including
Kautzsch I936, 202) originally stood at the extreme chariot racing) and venationes (involving bulls and wild­
southwest of this area before being moved to its centre in animals). These probably took place in some open space
the l 920's (cf. Akok I955, fig. 2). Moreover, on the south close to the city, enclosed and provided with seating on a
side of this space, excavations in 1995-96 revealed the temporary basis for the events, presumably the Ancyran
back wall of a building at least 31 m long, which evidently locality called 'Campus' in late Roman sources ( V. Plat.
faced north (Temiszoy et al, 1996). While its precise date 425 (ed. Migne)). We might assume it was located either
remains uncertain, both its scale and style suggest it on the level ground directly west of Ancyra, now occupied
belonged to a substantial structure, conceivably a stoa. by the Gern;:lik Park1 and the (perhaps appropriately
As for the 'Temple of Augustus' itself, the limited named) Hippodrom, or immediately east of Ulusdag1, the
excavations of the early I930's failed to reveal conclusive area now occupied by the Cc11tral Dolmu� Station.
evidence regarding its date or original form (Krencker These amenities of early Tiberian date apart, Ancyra
and Schede 1935, passim). The modern consensus is that gradually acquired the other usual standard features of a
it was built during the final years of Augustus' reign, classical city, such as a theatre, at least two bath-houses,
perhaps initially as an Ionic tetrastyle temple, measuring and an aqueduct. The first of these was discovered and
13 x 30111, later transformed by adding a Corinthian excavated in I982, and has been provisionally dated to
octostyle pseudo-dipteral colonnade and steps to form a the Ist century AC (Bayburtluoglu I986; this paper,
structure some 42 x 55111 (Guterbock 1989, 156; cf. Cooke Figure 1.5). It lies at the foot of the Kaledag1, the ima
1998, 26-7; this paper, Figure 1.3). Many believe it was eavea and central section of the summa eavea being
from the first intended as a ceremonial centre for the carved from the bedrock, the remainder built of andesite,
Galatian imperial cult, for the eel/a walls were re-cut to with local 'marble' used for decorative details. The ima
inscribe the text of Augustus' Res Gestae, and one anta eavea was divided into four eunei by three sea/aria, with
carries a list of the first 24 arehiereis, the first of whom a dia::oma at a level corresponding to the I0th or I Ith
dates to e. AD 19/20 (Bosch 1967, no. 51; Mitchell 1986, row of seating, which presumably gave access to the
28-9; 1993a, I 08). Yet the temple is purely Hellenistic in tribunalia over the aditus maximi, with a further I5 or so
plan and style, unlike the 'official' Roman design of a 12 rows of seating in the summa eavea, giving an overall
podium with steps at the front only, as in such 'imperial' diameter for the theatre of e. 56111. The wooden-floored
buildings as the 'Temple to Augustus' at Pisidian Antioch, seaenaefi·ons is clearly part of the original structure, but
built e. AD 2 (Mitchell and Waelkens 1998, 167). Thus the proseaena is equally evidently a later addition of more
the possibility must be allowed that it was perhaps than one phase, perhaps replacing an earlier timber
originally dedicated to a local deity, probably Men, with version. Despite such indications of economy in con­
or without /'deter Theon, and was only later adapted for struction, the seaenaefi·ons was decorated with'marble'
use by the imperial cult (Tuchelt 1985, 317-19; Vannltoglu statuary, including a cloaked male and a standing Pudieita
1992). Such indeed appears to be confirmed by the'priest figure, although the only piece of architectural decoration
list', for Pylaemenes, the fourth arehiereus, presented land found was a voussoir in the form of a Silenus head. That
at Ancyra for use as a sebasteiun, a structure which surely apart, the entrances to the theatre call for some extra
cannot be anything other than the official centre for the comment. They take the form of conjoined itinera
imperial cult. There is no evidence for its location or versurae and aditus maximi, a plan apparently un­
appearance, although it need not have been anything more paralleled in Asia Minor, although comparable arrange-
6 Julian Bennett

487 -IX lJ

424

14
- - - .L -
I
HIPPODROM 1\\1.
---- .! __
! ·,
I '·--'
I I
I 6 I
I
----L --�- -� I I
--, -- I
I 1
I I 423
423 I I
I 7 I
- - - .L - - - - _,
I I
GEN<;'LiK PARK! I
I
- -I-- - I - - -1 - - - I - -
I I
I
I

422 422

488 489
487

Figure 1.3 Roman Ancyra, showing principal topographic .features, known sites of Roman date, and restored street
plan. Key: I) the Ulus Meydani site; 2) the 'Temple of Augustus'; 3) the Theatre; 4) the Askeri Cezevi site; 5) the
'Caraca/lan Baths'; 6) the Yeni Meclisi site; 7) the Ziraat Bankasi site; 8) the Ulus Belediye site; 9) the 'Halk Evisi' site.

ments are to be found at the Gerasa South Theatre (Segal 2nd century date is suggested by its construction method,
1995, fig. I 02). of alternating rows of bricks and andesite blocks, and the
Of the two bath-houses recorded at Ancyra, that found existence of the so-called 'Caracallan Baths', a much larger
at the Askeri Cezevi site in 1946 measured some 30 x bath-house built (probably) in the mid- or Iate-2nd century.
30111, and appears to have been of bi-axial type (Akok This latter complex was first investigated in 193 I , when
1955, 323-29, and fig. 9; this paper, Figure 1. 6). No roadworks in <;:ankmkap1 revealed an open area surrounded
detailed report is available for this structure, but an early by rooms, one of them containing a life-size bronze imago
 I Ancyra, Metropolis Provinciae Galatiae 7

The bath-house walls were built of alternating courses

of 4-rowed brick and andesite blocks, with local 'marble'
used for decorative details. The palaestra had 32 columns
IQ] [Q] [Q] [Q] [Q] [Q] [Q] IQ] on each side, each c. 6111 high with Corinthian capitals,
supporting, it seems, an inscribed architrave. The central
IQ] [QJ
room on the east side, where the bust of Trajan was
IQ] � [Q] [QI discovered, may have been a kaisersaal, the rooms to the
north presumably offices, while the paired rooms found
IQ]
on the north and south sides of the palaestra were perhaps
\ IQ] [QJ libraries and/or lecture halls. The main bathing complex
IQ] D [QJ
itself was fr onted by a range of at least two rooms, the
central one with a natatio, the north room, with its
\
\ IQ] [QI hypocaust, presumably an apodyterium. Behind this range
was a centrally-located tepidarium, with a plunge-pool
\ and flanking rooms, while the southwest part of the
r , •
\ [QJ complex was taken up by the caldarium with furnaces to
\ the rear. A number of service corridors, two of them
[QJ
\ underground, provided the necessary access routes for
r ,
[QJ maintenance, and there were indications that the structure
r ,
L, li was lavishly decorated with floor and vault mosaics,
\ marble veneering on the walls, sculptured friezes, which
li
r ,

b '
�: \Ii] Ii] Ii] included a figure playing a cithara, and statues ( cf.
\.
'\
.
/ .
�
,
/
/
Iii Dolunay 1941, pls. 84 and 85).
As indicated, the date of the complex is uncertain. It is
IQ] '\� , . /IQ] Ii] 161 Ii] Ii] Iii usually assigned to the early 3rd century, as the earliest
coins found with the structure belong to the reign of
Caracalla, and the method of construction is typical of this
/ \
period (cf. Dolunay 1941, 266; Foss 1977, 62 and 87;
I Arslan 1996; Dodge, 1987, 112). Some have even linked
I

[g]'·,___ ./ its construction with a visit to Ancyra by the emperor

'Marcus Aurelius Antoninus', presumably Caracalla on
his way east in 215 (Bosch 1967, no. 260; Mitchell 1977,
64-5; cf. Dio 77. 9. 6-7, on building work in the eastern
20 M provinces at this time). That apart, one Tiberius Julius
Drawn by B Claasz Coockson 2002 Justus Julianus, archiereus and ktistes ('founder') of the
Figure 1. 4 The 'Temple of Augustus': re-drawn from metropolis Ancyra, is commemorated on a series of
Krencker and Schede 1936. matching inscriptions stylistically dated to the 3rd century,
and erected by the phylai of Ancyra for his gift of a
balaneion, generally assumed to be this complex (Bosch
clipeata of Trajan (Dalman 1932, 121-32; cf. Bennett 1967, nos. 255-258: cf. Broughton 1938, 778; Erzen 1946,
200 I , 20 I and pl. 20). Originally identified as the '.forum' 98-9; Mitchell 1993a, 114-16 and 214). This may well
of Roman Ancyra, it was correctly identified as the be, but the dating of inscriptions on stylistic grounds alone
palaestra of a bath-house after fi.nther excavations in 1938- is an inexact science, while the Caracallan coins found at
41 (Ank 1937, 49-51; Dolunay 1941, 264-66; Akok 1968; the bath-house merely indicate it was standing by that
this paper, Figure 1.7). It belongs to a group of Anatolian period. An earlier date is possible, for architrave fr agments
bath-houses which have been classified as the 'Bath­ found in the 1931 work are decorated in 'Hadrianic' style,
gymnasium' type (Yegiil 1995, 251-313), and it perhaps and bear an inscription commemorating a structure
covered an area of about 160 x 200m, making it one of the dedicated to the metropolis by one Titus Cornelius,
largest bath-houses of its kind. If, that is, it was ever although these may derive fr om the nearby colonnaded
completed. Excavation of the extreme southeast of the road (Dalman 1932, 125; Bosch 1967, no. 145; Mitchell
complex in 1944 uncovered walls and rooms of a and McParlin 1995, no.35; cf. Cooke 1998, 47).
completely dissimilar plan to those on the northwest (Akok Although the 'Caracallan Baths' lie only some 200111
1955, 311-15 and fig. 3, and 1968; cf. this paper, Figure from the Ankara <;:ay, they are 45111 above this stream, and
I .7). The published report provides no information of date appear to have been supplied with water by Ancyra's
or the physical relationship between these remains and the aqueduct. The numerous pierced blocks of andesite found
bath-house proper, but fr om their description, it seems re-used in the early medieval defences on the Kaledag1
they might belong to the late- or even post-Roman period. and at several other locations west of here, including - it
8 .Julian Bennett

''
''
,, ''
/
,, ''
/
''
/ ''
/
''
/
/
/
-------- \
\
I '' \
I '' \

I
I '' \
\
I '' '' \
I \ \
' \
I
I
'
' '' \
\
\ \
\
I
I
\ \ \
\ \ I

I
I I \
\
\I \
I
I \ I \
I I I
I
I I \
I
I I I
I I
I
'
I
I
'
---- - --- I
I
I
I
I I
I - - 1- - - - - - - -

'- - , 1- - - , 1- -
1 I I I
I I
II I I I
I
: , _ _ _ __I :
1 _ _ _ _ _ _1

5M
Drawn by B Ctaasz Cocckson 2002

Figure I.5 The Theatre; re-drawn .fi·om Bayburtluoglu I986, f ig 2.

would appear - the'Caracallan Baths' - indicate it was of 1.6). Again, further comment is confounded by the lack of
inverted siphon type, the blocks having the usual female/ any detailed report or plans, as is likewise the case with
male sockets at either end, and some with a hole in the the building found at the Nurettin Ersoy Otel site, directly
upper surface for cleaning purposes. The origins, course east of the ' Caracallan Baths' in 1947 (Figure 1.3):
and date of the system are, however, unknown. Given that apparently abandoned and/or destroyed in the 3rd century,
these blocks were extensively used in the southeast it had at least one apsidal heated room, floors of opus
sections of the early medieval fortification on the signinum, and walls covered with frescoes and marble
Kaledag1, at an elevation of 980m, it is assumed that the veneer (Akok 1955, 315-22, and fig. 7). As for any other
aqueduct passed nearby, suggesting its source was structures of early Roman date in Ancyra, we have brief
probably near the headwaters of the Ankara <;:ay on the reports from the 1920s and 30s of remains of that period
slopes of Kure Dag1, some 30km distant. As for its date, on the south slope of the Kaledag1, and on the sites of the
all that can be said is that the earliest known inverted Yeni Meclisi, the Ziraat Bankasi, the Ulus Belediye and
siphon system in Asia Minor is thought to be that at Patara, the'Halk Evisi', now the National Museum of Art and
built in the Flavian period (Coulton 1987, 80). Sculpture (Ank 1937, 47-9; Akok 1955, fig. I; this paper,
Very little else can be said for certain concerning Figure 1.2). To these we can add the discovery of several
Ancyra during the early Roman period. The street found ' Roman' tombs and the remains of a further bath-house,
running south from the presumed agora in 1995-96 was decorated with mosaics, at the west end of Gen�lik Park1
lined with a series of buildings, apparently shops: their (Jerphanion 1926, 223; Ko�ay 1939, 61).
exact date has not yet been determined (Temiszoy et al, Physical remains apart, an inscription indicates that the
1996). Also of an unknown but probably early Roman po/is possessed the expected bouleuterion (Bosch 1967,
date are the structures found next the Askeri Cezevi bath­ no. 117; note the J): phylae, Hiera 811/aia, perhaps located
house, later converted into a single dwelling with a ' court­ in its vicinity). In addition, from the spectacu/ae recorded
yard' (Akok 1955, 327-28, and fig 9; this paper, Figure on the · priest-list', the tombstones of gladiators who
 I Anqra, Metropolis Provinciae Galatiae 9

..
��TjS ::�J" l( )

�i
20 M

r
Drawn by B Claasz Coockson :?OD:?

Figure 1.6 The Askeri Ce=evi bath-house; re-drawn.fi·om

Akok 1955, f ig 9. 1

perished in Ancyra, and the presence of a procurator

formerly in charge of gladiatorial shows, it has been
suggested that there may have been an amphitheatre (e.g. ,
50 M
cf. Bosch 1 967, nos. 1 49-52, 1 88-94 and 276; Mitchell
1 977: 72-5; Erzen 1 946, 97-8). This is most unlikely: Drawn by 8 Claasz Coockson 2002

such purpose-built structures are exceedingly rare in Asia Figure I. 7 The 'Cara callan Baths '; re-dra wn.fi·om Akok
Minor, only being known of at Pergamum, Cyzicus and 1955, fig 3.
Antiocheia. True, the theatres in many of the Anatolian
poleis were adapted or even specifically designed for such
shows (e.g., Aphrodisias, Myra and Xanthus), but that at temple presumably associated with one or other of these
Ancyra shows no evidence for such a modification: it is cults (Arslan 1 99 1 , no. I), if it was not that dedicated to
likely, therefore, that all such spectac11/ae took place at Apollo.
' Campus', within temporary structures of the type already
posited for the Tiberian period.
On the other hand, there is evidence to suggest that
there were at least two other temples in Roman Ancyra
Late Roman Ancyra
besides the'Temple of Augustus'. One was the Temple of It is uncertain if Ancyra suffered directly from the Gothic
Zeus recorded by Pausanias, presumably as either Zeus attacks of the 250s-260s: the area surely did, and the
Trapezeus or Zeus Taenos, which is shown on coins as a Goths were not loath to raid the undefended poleis of the
hexastyle structure (Paus. 1 .4.5; Arslan 1 99 1 , no. 4; cf. region (Mitchell 1 993a, 235-36). It certainly fell to Queen
the V phy/e, Dias Tra pe=on, and the XII, Dias Taenon, Zenobia's marauding Palmyrene army in 27 1 , however, lo
and Bosch 1 967, 2 1 I, a dedication to Zeus Taenos). A be recaptured by Aurelian later that same year (Zos. 50,
second was probably a Temple to Apollo, who seems to 1 -2). It is presumably to one or other of these incursions
have been the favoured deity at Ancyra in the early 4th that we might associate the destruction of the private
century ( J/ .Plat. 404-425 (ed. Migne); cf. also Mitchell building on the Nurettin Ersoy Otel site, and it perhaps
1 982b, 94). These apart, our sources indicate local cults also to this general period that we should ascribe an
to some 1 8 other classical deities, ranging from Athena to incomplete but evidently late Roman inscription which
Victoria, of whom Artemis and Athena also seem to have refers to the construction of civic defences. It com­
played a prominent part in the religious life of Ancyra memorates an anonymous benefactor, who 'during the
(Arslan 1 99 1 ; Bosch 1 967, passim; Mitchell 1 982b, 94). time of famine and barbarian attacks' built the defences of
Although it is unlikely that each had their own formal the po/is ' from the foundations to the battlements' , as well
temple, an Ancyran coin of Nero depicts a tetrastyle as restoring two ' ruined' buildings, the 'gymnasium of
10 ./11/ian Bennett

Polyeidon · and the office of the 8011/ographoi, the last western Empire (Opitz 1934, no. 20). Various accounts
having been disused for some period of time (Bosch 1967, confirm that the po/is was a thriving place in the mid-4th
no. 289). A likely context for such building activity would century AC, and refer for example to such buildings as a
be in the aftermath of the Palmyrene attack, and these new church dedicated in 358 (Sozomen 4.13). Indeed, if
defences may well be represented by the 12111 high section we are to believe Libanius, Ancyra was radically trans­
of walling with three projecting 4m wide square towers formed in the 3 60s, as he singles out one Maximus,
recently revealed immediately west of the 'Temple of governor in 362-364, for his substantial contributions to
Augustus'. Built mainly of andesite blocks, with alter­ the urban appearance of the po/is, contributions which
nating tile-courses in the towers, it contains several reused evidently included a variety of public buildings as well as
column shafts and at least one homos, but no other fountains and nymphaea (Lib. Ep. 1230). Indeed, it is
architectural material or tombstones. In its general tempting to identify M aximus with the unknown governor
character and evidently limited perimeter - it excludes the commemorated sometime in the late Roman period for
area between Ulusda I and the 'Caracallan Baths' - it building a wall to 'ensure the safety' of the po/is (Bosch
closely resembles the walls erected at several Gallic 1967, no. 290), although this person may be the same man
civitates at this time, as for example at Amiens and Bavai. noted earlier, as having built fortifications at Ancyra
Nonetheless, despite the apparent availability of money, apparently in the aftermath of the Palmyrene attack. After
manpower and materials for reconstruction work at Ancyra all, given that Ancyra is one of only six urban centres
in the later-3rd century, literary sources suggest that the shown with fortifications on the Tahu/a Peufingeriana,
metropolis and its immediate region took some time to we might be justified in assuming that sometime in the
recover from the effects of these two attacks. Just before 3rd-4th century the po/is was given imperial assistance to
the end of Aurelian's reign, for example, a grain merchant build a defensive circuit which included a more substantial
by name of Philumenus brought cereals to Ancyra from area than that indicated by the defences surviving west of
Lycaonia, suggesting that supplies were not available the'Temple of Augustus'. If so, however, there is as yet
locally: unfortunately for Philumenus, his reward was no indication as to where these defences were.
denouncement and execution for his Christian beliefs It is perhaps to this general period that we should also
(Syn.Eccl. Const. 263 -264 (ed. Delehaye)). Indeed, it is ascribe the hypogea found at the west end of Gern;:lik
possible that the apparently dire situation did not improve Park1 in the 1930s (Akok and Pen�e 1941). Likewise,
to any great extent until towards the end of the 3rd century, another undated inscription from Ancyra, which com­
as the Life of St. Clement of Ancyra indicates famine and memorates the work of Joannes, son of Eutychikos and
great mortality in the region in c. 283 ( V. Clem. 816-893 anatellon ('refounder (of the po/is')). He'restored' many
(ed. Migne)). structures, including inter alia the'hall of Polyeidon', the
We might assume, therefore, that it was not until the 'building of Theodotus', the aqueduct and water distri­
reforms of Diocletian that the overall situation improved bution system, and the prison; re-roofed an unnamed
in Galatia in general, and in Ancyra in particular. Now the public building and the 'palatium' (sc, the praetorium?);
metropolis of a reduced polity, important information and'marbalised' a second unnamed edifice (Bosch 1967,
regarding Ancyra's appearance and topography at the time no. 306). His name indicates a post-Constantinian date,
is supplied by the narratives of three martyrs who suffered and we might associate his activity with Ancyra's evident
there during the Great Persecution. From the Life of St. revival in the mid-late 4th century.
Plato, for example, executed under Galerius, we learn that Literary accounts remain our principal source for
he was tried in the 'basilica' opposite the 'Temple of Ancyra in 'Late Antiquity', the period from c. 395-620.
Zeus', and executed at the 'Campus' ( V. Plat. 404-425 We might note, to begin with, that Ancyra was Arcadius'
(ed. Migne)). The afore-mentioned St. Clement, on the favoured summer resort at the turn of the 4th century AC,
other hand, perhaps martyred under the same emperor, which undoubtedly indicates that its post-Diocletianic
was tried, executed and buried at an Ancyran locality revival had made it a suitable place to conduct imperial
called 'C,J,ptus', where a martyrium evidently already business and accommodate his retinue (Foss 1977, 50-1).
existed at the time ( V. Clem. 889-892 ed. Migne). As for Certainly, by the end of the 430s, there were a number of
the Life of St. Theodotus, apparently martyred in 312, this ecclesiastical premises in or adjacent to the po/is, including
refers inter alia to two martyria close to the po/is, the a cathedral (Pall. Hist. Laus. 67; Nilus 968-1060, passim),
governor's 'pra etorium' - seemingly located in the centre a church dedicated to St. Plato, apparently Ancyra's patron
- a fountain, shops and private houses (Mitchell 1982b, saint (Nilus Ep. 1 1.178.291), a church reserved for use by
104-5). the Novantian sect (Sozomen 8.1), a monastery on the
Other Christiological sources certainly suggest that the Tamerlanedag1 (Nilus Or. Alh. 79 (ed. Migne)), the convent
civic fortunes of Ancyra had radically improved by the of Magna, and a hospital and 'hospice' (Pall. Hist. Lazts.
early 4th century. In 324/325, for example, Constantine 66). To these we might add the conversion of the'Temple
apparently made Ancyra his initial choice for the First of Augustus' for use as a monastic church, probably in this
Ecumenical Council, although in the event Nicaea was general period (Foss 1977, 65-6).
chosen for its convenience of access for the bishops of the Christiological sources confirm that Ancyra remained
 I Ancyra , M etropo l i s Pro v i nciae Ga latiae 11

an active and prosperous centre dur i ng the later 5th, 6th Bibliography
and early 7th centuries, despite outbreaks of pesti lence
Akok. M . . 1 95 5 . Ankara Sehri ir;inde rastlanan i l kr;ag yerl e�­
and famine ( Foss 1 97 7 , 54-60). In 6 1 1 -6 1 3 , however, the
mesi nden baz1 zlen ve li<; a ra�t1rma yeri. Belletin 1 9. 3 0 9-29
i m med i ate region fe l l prey to the Pers i a n s , a lthough
Akok. M .. 1 96 8 . A n kara S e h r i n d e k i Roma 1-l a m a m 1 . Tiirk
A ncyra itse l f does not seem to have been occupied and Arkeoloji Dergisi 1 7. 5-3 7 .
they were soon forced to withdraw (Foss 1 97 5 , 722-2 3 ) . A kok. M . and Penr;e. N . . 1 94 1 . An kara i stanyonuda B u l tman
The strateg i c s i g n i fi c a n c e of A n c y ra for c o ntro l o f B i zans Devri M ezan n m Nakh. Belle/in 5. 6 1 7-22.
Anato l i a w a s , on t h e other h a n d , c l early recogn ised b y the Arik. R . . 1 93 7 . Les Res u l tats des fou i l les faites a Ankara par la
invaders : when they returned i n 622, they c aptured the societe d " h i stoire turque. la Turquie Kemaliste 2 1 /22. 47-56.
p l ace, and k i l led or enslaved its c i tizens ( Foss 1 977, 70- Arslan. M . . 1 99 1 . ·The Coi nage o f Ancyra in the Roman Peri od · .
1 ). V i v i d evidence for the attack may have been d i s­ i n C. S . Lightfoot ( ed ) . Recent Turkish Coin Hoards and
covered during the excavation of the ' C aracal lan B ath s ' , Numismatic Studies ( Ox ford 1 99 1 ). 3-42.
Arslan. M . . 1 996. Greek and Greek I mperial Coins Found during
in the form of thick layers of bui l d ing debris a n d ash,
the <;:ankmkap1 Excavations at An kara. I n R. Ashton ( ed . ).
a s s o c i ated with c o i n s o f H e rac l i u s a n d ' S a s s a n i a n '
Studies in A ncient Coinage .f r om Turkey. 1 07- 1 4 ( London ).
artefacts ( Foss 1 977, 7 1 ) . Bayburt l uoglu. i.. 1 98 6 . A n kara Antik Tiyatro s u · . A nadolu
So, it seems, ended t h e l i fe of Ancyra, metropolis 1'/eden(vetleri Mii::esi - 1 986 J '11!1g1. 3 9-43 (Ankara ) .
provinciae Ga/atiae, for after recapture in the m id-7th Barnett. R. D . . 1 974. T h e Euro pean M erchants i n Angora A nal
century, a l l occupation was centred on the newly-bu i lt Stud 24. 1 3 5-42 .
Thematic fortress on the Kaledag1, and the wal led area Bennett. J . . 2 0 0 I . Trojan Optimus Princeps ( London ).
occupying its western s lopes. Th is, the ' famous and great Bosch. E.. 1 967. Quellen ::ur Geschichte der Stadt A nkara im
castle, the powerfu l and fortified Ancyra ' of the early A ltertum ( An kara ) .
medieval period (Digenes Akrites 9- 1 1 ), was to change B roughton. T . R. S . . Roman Asia. I n T . Frank ( ed . ) . .-I n Economic
hands many times over the next 4 5 0 years, unt i l its final Survey of A ncient Rome 3. 499-9 1 6 ( Balti more ) .
Cooke. S . D . . 1 99 8 . The Monuments o,(Roman .-l ncyra Reviewed
capture by the Se lcuks i n 1 1 2 5 , when occupation spread
( un p u b l ished MA D i ssertation. B i l kent O n i vers i tesi. An­
to the south of the Kaledag 1 . Another 250 years were to
kara ) .
pass, however, before the U lusdag1 was agai n reoccupied, Cou lton. J . J . . 1 98 7 . Roman Aq ueducts i n Asia M i n o r. I n S .
when the Ottom ans bu i l t a mosque a n d tu rbe com­ Macready a n d F . 1-1 . Thompson. ( ed s . ). Roman A rchitecture
memorating Hac1 Bayram, Angora ' s favourite ' sa i nt ' , next in the Greek /Farid. 72-82 ( London ) .
to the upstanding eel/a of the ' Temple of Augustus ' , which Dal man. 0 . . 1 93 2 . · 1 93 1 · d e A nkarada M eydana <;: 1 kan l an Asan
was now converted into a medrese. I t s t i m u l ated the A t i ka · Tiirk Tan/11. A rkeoloji ve Etnogrq[11a Dergisi I .
regeneration of U l usdag1 as the re l igious and j u d i c i a l 1 2 1 -3 3 .
centre of Angora, a process forma l ised, in October 1 92 3 , Dodge. 1-1 . • 1 98 7 . Brick Construction i n Roman Greece and Asia
when Atati.irk chose the p l ace as the capital fo r the N ew M i nor. I n S . Macready. and F . 1-1 . Thompson. ( eds. ) Roman
Repu b l i c of Turkey, and U lusdag1 for h i s seat of govern­ Architecture in the Greek /Farid. 1 06- 1 6 ( London ) .
ment. The process of o b l i terating c l assical Ancyra now D o l unay. N . . 1 94 1 . Tlirk T a n h K u ru m u Ad m a Yap 1 lan <;:an­
kmkap1 1-l oyligli 1-la lhyat1 . Belletin 5. 26 1 -276 ( An kara ) .
began i n earnest, a lthough Atati.irk h i mself was exceed­
Erzen. A.. 1 946. i!kcagda Ankara ( A n kara ) .
ingly keen to preserve its past (cf. G uterbock 1 9 89). We Eyice. S . . 1 97 1 . Ankara · n 1 11 E s k i B i r Resm i . A tatiirk Ko,?fer­
can only hope that more e n l ightened attitudes w i l l aga i n anslar l II ', 1 9 7). 6 1 - 1 24.
s o o n prevai l before what remains is fi n a l ly lost. Foss. C.. 1 97 5 . The Persians i n Asia M i nor and the End o r
Anti q u i ty. Eng /-list Rev 90. 72 1 -47.
Foss. C . . 1 977 Late Antique and Byzantine A nkara DOP 3 1 .
29-8 7.
Ack11owledgeme11ts French. D . . 1 972. A S ixteenth Century Engl ish Wool M erchant
I am most p l eased to acknowledge my gratitude to Susan i n An kara? . .-I na/ Stud 22. 24 1 -48.
Ugurlu for perm i s s i o n to u s e h e r unpu b l i s h e d M . A . G uterbock. 1-1 . G . . 1 989. The Temple o f Augustus i n the 1 93 0s.
I n K. E mre. B. 1-l ro uda. M. Mel l i n k and N. Ozgli<; ( eds. ).
research i n preparing this paper (Cooke 1 99 8 ) : fo r the
.-l natolia and the Near East Studies in Honor o,(Ta!jtn 6::giir;.
overa l l context of Ancyra in the pre- and Roman periods,
1 5 5-5 7 (Ankara ) .
M itche l l 1 99 3 a and b, passim, and for Ancyra i n the late H a l ey. S . . 2000 Caveat Emptor: the intellectual consequences
Roman period, Foss 1 97 7 . I wou l d also l i ke to thank a,( undocumented excavation, with special reference lo
tN orbert Karg for h e l p ing w ith p h i l o lo g i c a l matters ; Roman period archaeological material .f r om Turkey ( un ­
Barbara Helwing, for assi stance w ith G erman sources; p u b l i shed M A D i ssertation. B i l kent O n i vers i tesi. Ankara ) .
and J acques Morin, Ben C l aasz Coockson, Cadoc Leigh­ i nan. J . . 1 994. Boubon Sebasteionu v e /-/eykelleri U::erine son
ton and Ay�e Be lgen, who respectively helped with the Ara,v tmnalar ( I stanb u l ) .
G reek inscriptions, pro v i ded the i l lustrations, c larified Jerphan ion. G . d e . 1 926. 1\/elanges c/ '.-lrcheologie .-l natolienne
Christiological references, and procured the base maps . I ( Be i ru t ) .
a l one, however, bear the respon s i b i l ity for the inter­ Jones. A. 1-1 . M . . 1 940. The Greek City ( Oxford ) .
Kautzsch. R . . 1 93 6 . Kapitellstudien ( Lei pzig).
pretations and views expressed in th is paper.
12 Julian Bennett

Ko�ay. 1-1 .Z .. 1 93 9 . The Strata of C i v i l isation in Ankara, La Emperors fro m Inscriptions and Papyri ( MA MA 1 7 8 :
Turquie Kemaliste 3 1 , 5 8-64. Phi ladel phia).
Krencker, D., and Schede, M . , 1 93 6 . Der Tempel in A nkara Opitz, 1-1 . G., 1 934. A thanasius Werke I I I : I ( Berl in).
(Berl i n ) . Ozgli9, T., 1 946. Anadolu Ara�ttrmalan. Belletin 1 0, 5 5 7-97.
M e t i n , M . and A ka l m , M, 1 999. Ankara-U l u s Kaz1 s 1 Frig Ozgli9, T. and Akok, M., 1 95 5 . A11 1 t-kab1 r Alanmda Yapt lan
Sera111 1g1, Anadolu Medeniyetleri Miizesi - 1 998 Yllhg1, 1 4 1 - TlimUIUs Kazt lan, Belletin 1 9, 2 7-56.
62 (Ankara). Remy, B . . 1 989. Les Carrieres Senatoriales Dans Les Provinces
M itchell, S . , 1 977. R(egional) E(pigraphic) C(atalogues) A(sia) Romaines D 'A natolie Au Haut-Empire ( I stanbul).
M ( i nor), Anal Stud 21, 63- 1 03 . Robert, L . , 1 960. I nscri ption Ago n i stique d ' Ancyra . I n L.
M i tchell, S . , 1 982a. Regional Epigraphical Catalogues of Asia Robert, Hel/enica 1 1 / 1 2, 3 5 0-68 ( Paris).
Minor 11 The A nkara District, BAR Int Ser 1 35 (Oxford ) . Segal, A . , 1 995. Theatres in Roman Palestine and Provincia
M itchel l, S . , I 9 8 2 b . T h e Life of Saint Theodotus o f Ancyra. Arabia ( Leiden ).
A nal Stud 32, 93-1 1 4. Temiszoy, I . , Arslan. M . . Akalm, M . and Metin. M . . 1 996. Ulus
Mitchel l , S . , 1 986. Galatia Under Tiberius, Chiron 1 6, 1 7-3 3 . Kaz1s1 1 995, Anadolu Medeniyetleri Miizesi - 1 995 Yil/1g1,
M itchell, S . , 1 993a. A natolia: land, men and gods in Asia Minor 7-3 6 (Ankara).
/: The Cells and the Impact of Roman Rule ( Oxford ) . Tezcan, B . . 1 964. Yal111cak I 'ii/age Excavation in 1 962-63
M i tchell. S ., 1 993b. A natolia: land, men and gods i n Asia Minor ( Ankara).
11: The Rise of the Church (Oxford ). Tezcan, B . , 1 966. Yal111cak I 'ii/age Excavation in I 96-1 (An kara).
M itchel l, S . and McParl in, P . , 1 99 5 . Greek I nscri ptions in Tuchelt. K . , 1 98 5 . Zur l denti tatsfrage Des 'Augustus-Tempels'
Ankara, ( unpubl ished catal ogue in the British Institute o f in Ankara · AA. 3 1 7-22.
Archaeology, Ankara). Varm l toglu. E . . 1 992. Meter Theon ' A nadolu Medeniyetleri
M itchell, S . and Waelkens, M . , 1 998. Pisidian A ntioch: the site Miizesi - 1 99 I Yil/1g1, 3 9-43 (Ankara ).
and its m onuments ( London). Yegli l , F . , 1 99 5 . Baths an d Bathing i n Classical A ntiquity
Moretti. L . . 1 95 3 . Iscrizzioni agonistische greche ( Rome). ( London ).
O l iver, J . 1-1 . . 1 989. Greek Constitutions of the Early Roman