176

Antonio Invemizzi

THE CULTURE OF PARTHIAN NISA
ensemble,
its

177

The corpus of rhytons

itself

appears to be a

lively

diversity within

an

INVERNIZZI,

apparent homogeneity of function coming from the participation of a multiplicity of
hands, which, however, raises the question of the chronological homogeneity of the
group.

Nisa', in A. 2001c, 'Musicanti nei fregi del rhyta di Iranica 35), Louvain, 569-85. Guillemin (Acta

Monumentum

Marcelle Duchesne-

Above all, it is perhaps these figurative artefacts which, on account of the wealth

of their iconography, emphasize the problem of the relationship with known centres of
Greek-style artistic production in the provinces within the dissolved Seleucid empire.

As one,

scholars have recognized the originality of their artistic character, substantially

1977. Rodina Parfjan, Moscow. nord della Sala Rotonda. Rapporto prehmmare LIPPOLIS, C. 2002. 'Nisa-Mithradatkert: Fedificio a 2000-2001', Parthica 4, 47-62. delle campagne di scavo 55-65. 2003. 'Review of Pilipko 2001', Parthica 5, 1 1 TPPOLIS, C. della Sala Rotonda. Rapporto prehmmare C. 2004. 'Nisa-Mithradatkert: fedificio a nord LIPPOLIS, 161-77. delle campagne di scavo 2002-2003', Parthica 6,

KOSHELENKO, G. A.

.

extraneous to Mediterranean culture, but they have pointed to different routes for their

MASSON

M

E
E.

& PUGACHENKOVA, G. A.
illjustracy,

1956.

P«tf«^

West and Babylonia. And yet, despite the difficulty in forming judgements, due to the extreme rarity of comparable evidence in the Seleucid empire and in Hellenized Asia, the most natural solution may be closer at hand. The recognition that there were Greek-oriented artistic workshops active in Nisa in the Hellenistic period, may suggest that these spectacular artefacts were part of a local production. Arsacid Parthia may well aspire to be recognized as the
relationships, towards the East

and

turkmenskogo naroda, Al'bom

India, or towards the

MASSON, M.

& PUGACHENKOVA, G
Florence.
S.

Trudy YuTAKE, Ashkhabad. A. 1959. Parfjanskie ntony Nisy, Trudy
t

YuTAKE

IV,
,.

Moscow, Ashkhabad.

MASSON, M. E. & PUGACHENKOVA,
Mesopotamia
1),

, Ti (Monografie di G. A. 1982. The Parthian Rhytons of Nisa
. ,

_

„_

MELIKIAN-CHIRVANI, A.

1996. 'The Iranian

wine horn from Pre-Achaememd Antiquity to the

seat of

one of the most important schools of Greek or Greek-trained artists Hellenistic Asia, alongside Seleucid Babylonia, Elam, Bactria and Gandhara.

85-139. Saf&vid Age', Bulletin of the Asia Institute 10, trom via 2000. 'The manufacturing technique of the rhytons MKRTYCHEV, T. TREINER, U.

&

in

Nisa', Parthica 2, 55-67.

PFROMMER, M.
PILIPKO V N.

Catalogue of the Collections. J he 1993. Metalwork from the Hellenized East.

,

..

J.

Paul Getty Museum, Malibu.
1996. Staraja Nisa. Zdanie s
.

Kvadratnym Zalom, Moscow.
period,

izuchenija v sovetslcy PILIPKO' V. N. 2001 Osnovnye itogi arheologicheskogo Turkmenistanapory rabovlademja A. 1958. Puti razvitija arhitektury juzhnogo PUGACHENKOVA,

Moscow

G

BADER,

A. N.

& USUPOV,

Invernizzi (ed.), In the

Kh. 1995. 'Gold Earrings from Northwest Turkmenistan'. Tn A. Land of the Gryphons (Monografie di Mesopotamia V), Florence, 23-38.
I.

BERNARD, P. BERNARD, P.
1991, 31-8.

1985. 'Les rhytons de Nisa.

Poetesses grecques', Journal ties Savants, 25-1 18.

1991. 'Les rhytons de Nisa; a quoi, a qui ont-ils servi?' In Bernard

&

ifeodalizma, Trudy YuTAKE VI, Moscow. , . najdennyh SERGEEVA, T. V. 1972. 'K voprosu o meste izgotovlenija ritonov, narodov Vostoka, 5, 66-75. Nisa', Gosudarstvenny Muzej islatsstva empire des Arsacides (Acta Iranica 32), Louvain. J. 1993.
.

A *, c+^i. na gorodiscn Staraja
-

WOLSKI,

V

Grenet (eds)

BERNARD, P. & GRENET,
ecrites et

F. (eds)

1991. Histoire et cultes de VAsie Centrale pre-islamique. Sources
Paris.

documents archeologiques,

CHUVIN,

P.

1991. 'Fetes grecques sur les rhytons de Nisa'. In Bernard

& Grenet (eds)

1991, 23-9.

INVERNIZZI, A. 1989. 'Heracles a Seleucie du Tigre', Revue Archeologique 1, 65-113. INVERNIZZI, A. 1994. 'Die hellenistischen Grundlagen der fruhparthischen Kunst',
(1996), 191-203.

AMI

27

INVERNIZZI, A.
33-8.

1996. 'Old Nisa

and the Art of the Steppes', Bulletin of the Asia Institute 10 (1998),

INVERNIZZI, A.

1998. 'Parthian Nisa,

new

lines

of research'. In

J.

Wiesehofer

(ed.),

Das

Partherreich unci seine Zeugnisse, Beitrage des Internationalen Colloquiums, Eutin (27.-30. Juni

1996) (Historia Enzelschriften 122), Stuttgart, 45-59.

INVERNIZZI, A. INVERNIZZI, A.

1999. Sculture. di metallo

da Nisa. Culiura greca e cultura iranica

in Partia

{Acta

Iranica III serie, XXI),

Lou vain.
Orientis. Archaeological

2000. 'The Square House at Old Nisa'. In In altera parte Research in Pre-Islamic Central Asia, Parthica 2, 13-53.

INVERNIZZI, A. 2001a. Arsacid Palaces'. In The Royal Palace Institution in the First Millennium bc:
Danish
Regional Development and Cultural Interchange between East and West (Monographs of the Institute at Athens 4), Aarhus, 295-315.
2001b. Arsacid Dynastic Art', Parthica 3, 133-57.

INVERNIZZI, A.

174

Antonio Invernizzi
the Arsacids

THE CULTURE OF PARTHIAN NTSA
and
their

175

commissions may have attracted

artists

from Greek centres near

far to Nisa, with the express task of contributing to the preparation of the monumental rooms. In any case, royal patronage turned Nisa into a workshop in constant activity, fitting the economic importance of this genuine royal foundation in Parthia,

and

comparable to that which temples had in the Ancient Near East. These commissions certainly spanned the entire Hellenistic era, that is, the centuries
before Christ,

perhaps then

progressively

to

decrease with successive dynastic

upheavals, though the economic power of the royal foundation remained intact

throughout the Parthian age. Recent research indicates an important evolution with various phases of construction in the area of the Square Hall, whose clay sculptures are
probably
later

than those in the

Round

Hall.

The

virtuoso drapery and idealized but

lively and enquiring features of Mithradates (Fig. 10), almost like those of a philosopher, give way to the fixed expressions of the heads of warriors and princes in the

Square Hall

(Fig. 11).

Figure 8.

Old Nisa, Aphrodite Anadyomene.

Figure 9.

Old Nisa, Aphrodite Anadyomene.

must be recognized that both sculptors drew on a them in the most appropriate way into their figurative discourse, which is wholly original. Details aside, it seems unlikely that these marble sculptures were of Mediterranean origin, if for no other reason than the tricky problems of transportation posed by a sculpture like the Anadyomene. They are more likely to have been produced by Greek sculptors at Nisa itself, where, at about the same time, Greek sculptors, or Greektrained clay artists, were certainly also at work in the Round Hall. The creators of these marble figures were artists whose Western origins are called into question by the structure and modelling of the face of Aphrodite. The head carved at Susa by the sculptor Antiochus, which would not look out of place beside the sculptures commissioned by the Arsacids, the bronze Hercules of Mesene and the few fragments of sculpture at Seleucia on the Tigris, may be placed alongside the artefacts from Bactria to testify to the existence of a flourishing and varied artistic activity in the Greek cities and in the cities of Hellenized Asia. These works may have a parallel in the work of the Hellenistic sculptors at Nisa. If there were no workshops already, the new imperial ambitions of
it

of the Venus of Milos, Therefore,

common

repertoire of stylistic formulae, inserting

Figure 10.

Old Nisa, Round
Mithradates

Hall, clay head of
I.

Figure 11.

Old Nisa, Square

Hall,

day head.

172

Antonio Invemizzi

THE CULTURE OF PARTHIAN NISA
The

173

motif, as in the other Hellenistic subjects of the metalworker's art found at Nisa.
traditional familiarity

of the northern Greeks with the steppe-dwellers may, in various ways, have been rekindled in Arsacid Parthia.

The
tainly

routes taken by the Arsacid peoples to

come down

to Parthia, therefore,

may
cer-

well have remained open. Their importance, like that of the

nomadic element

from notable exceptions of the calibre of Tillya Tepe may have been greater than generally thought. But the possibility of a northern orientation in taste shown in high quality Parthian metalwork does not, of course, mean that most of the Hellenistic influences could not have arrived by other routes, the more classical ones, the more frequently used and also better researched on
not easily
verified in the settled areas apart

the ground. It

is

in fact along the routes that directly connect the

Mediterranean to the

Iranian world that the

movement of

decisive influences

on the manifestations of
the other sovereigns, the
fit

Arsacid kingship take place.

On

taking their place

among

Arsacids could not find in their ancestral traditions a language
status. So, to give expression to the celebration
ily

to express their

new
figure 7.

of this

new regal status

they drew heav-

Old Nisa, metop

he Seleucid anchor.

upon the language of the iconography and Greek

art that the Seleucid

empire had
well have been different routes followed by other motifs and influences the seen in the works of art recorded at Nisa. The most representative among these are marble statues which from the moment of their discovery were believed to be imported

already confirmed as a lingua franca.

There

may

The importance of great Hellenistic art in the figurative expression of the Parthian is well known. It was the Arsacids themselves who chose, among others, the epithet Philhellene, which certain kings added to their titles, and their Philhellenism
empire
is

always considered a fundamental feature of their politics and culture. In particular
is

as far as the art of the first Arsacids
at Nisa, the contribution

concerned, which

may be judged from

the finds

of Hellenic thought and Formengut to the expression of Arsacid kingship appears of paramount importance. But when we ask ourselves what
routes these influences followed, our almost total ignorance of the culture of Seleucid

from the Mediterranean. There is indeed no doubt that they were produced by Hellenic unbaked clay from artists of great talent. But the fragments of monumental statues of masthe Round Room, which were certainly made in loco, prove that their creators had protered Hellenistic styles of execution, making the possibility that there was a local unlikely. The sculptors who produced them were duction of works in marble far from
perfectly well aware of the tendencies

Parthyene places serious obstacles in the path of our attempt to understand the relationship between the events of the time and the formation of the empire. The existence

and development of the artistic production of the Mediterranean. There was certainly more than one of them, and their works belong to paralleled in different trends. The stylistic variety of the Nisa sculptures, moreover, is
Bactria,
level,

of a powerful state under Andragoras before the reign of Arsaces, however, and the possibility that his palace was at the centre of a movement of ideas, may lead us to conclude that the Arsacids, seizing power in a period characterized by independence

where the sculptures at Ai Khanum, which are at a clearly inferior qualitative also indicate the contemporary existence of different tastes in the everyday pro-

movements in other regions, too for example, in neighbouring Bactria entered into an ideological process which was already well under way and so acquired in loco their heritage of Macedonian and Seleucid conceptions of kingship. With regard to the forms in which this ideology manifested itself, too, it is permissible to think that

duction of the local workshops.
In particular with the Aphrodite Anadyomene (Figs 8-9), who the Arsacides probdrapery ably interpreted as their goddess Anahita, we see that the way in which the when seen in profile, the finds precise parallels in Aegean sculpture: especially

hangs

they developed their principal figurative elements directly from the
in loco.

design of the pleats in the fabric which fold around the right leg has a close parallel with one of the most famous of Greek sculptures: the Venus of Milos. But the parallel

Seleucid repertoire

The anchor
is

that appears

among

the

most important sym-

can be taken no

bols of Arsacid kingship,

clearly the

most

exquisitely Seleucid in origin (Fig. 7).

Alongside symbols of the divine, such as the club of Hercules and the astral figures of disc and crescent, the rosette (solar symbol of life), the lion's head (the kingly attribute

because the formal language of the more fragmented folds and the pictorial contrast between the torso and the drapery, heightened by the use of two different kinds of stone, gives a very different impression of the goddess. A glance at majesty the shape and structure of the face of this young goddess, so unlike the classic
further,

par excellence), the quiver on one hand and the Seleucid anchor on the

other,

may have

of the Venus of Milos,

is

enough to show both the
it is

originality

of the Nisa

artist's cre-

been chosen to indicate the characteristic inherent to the legitimacy of the new dynasty.

ation, as well as to declare

substantially independent of the Mediterranean context

170

Antonio Invernizzi

THE CULTURE OF PARTHIAN NISA
may be many
it is

171

bears to the existential concepts of the Arsacids, because there
for the presence of this object in the royal treasury at Nisa,

reasons

and

not certain that
suffi-

they themselves commissioned

it.

The

rich burial

mounds

at Tillya

Tepe provide

extent, The importance of their cultural background thus appears to have, to some the Arsacid empire, and it is even possible that influenced the new direction taken by steppes. The discovery at Olbia Hellenistic cultural influences reached Parthia via the

open to the peoples and customs of the steppes. On the other hand, judging from what remains of Nisa, their royal art appears to have been influenced above all by Hellenic figurative tastes.

cient evidence that the settled lands of Central Asia were

and and form of the art of the steppes but with its deepest and most spiritual concepts, though now expressed in a new figurative language and composition. A careful reading, in fact, makes particularly evident
still

However, another artefact reveals signs that ancestral concepts were

alive

is

representative of a direct link not with the taste

majesty, belonging to a rhyof fragments of an ivory frieze depicting a Parthian king in direct communication with ton that we may imagine to be like those at Nisa, proves that without using the Mediterranean, either via the these northern regions probably existed artefact typical of Parthian jewsouth or north of the Caspian Sea. The inclusion of an of an Eros, in the burial gifts of a grave in ellery, such as the gold earring in the form Caspian Sea, may region south of Lake Aral, between the lake and the

the

Uzboy

m

the ideology behind the iconography of the great ceremonial shield, part of whose applied metal decoration has come to light, because it is likely that in this case it really
is

a royal Arsacid shield (Fig.

5).

In fact, beneath the form and the

style,

which have

Hellenic references, both the subjects of the decorations as well as their composition,
in which palmettes alternate with eagles to

form a border around the central tridentshaped emblem, embody typically Scythian concepts. The three subjects must clearly be
interpreted as symbolizing respectively earth, air

south-east to north-west direction taken fact indicate one of these possible routes. different from their place of by these goods, whose place of manufacture is certainly north-west (but also north-east) discovery, finds a mirror image in the movement from of the few metal figurines discovto south-east suggested by the formal characteristics (the griffon, the aforementioned stag) in fact ered in the Square House. Some of these Central Asiatic areas, but significant connections with the figurative art of other

The

show

and water: that is, the tripartite kingdom of Scythian cosmology. The kings, therefore, could appear on parade protected by a shield whose representation of Scythian conceptions of the world through the language of symbols came to protect the person of the sovereign himself.

Hellenistic iconography with a others (Athena, Eros, the centaur-archer) reproduce Greece rather than that of the particular taste that reflects the culture of northern Asia itself. The silver harpy (Fig. 6), in parMediterranean, or points towards Central of a ceramic was probably one of the three feet of a pyxis, the precious version
ticular,

the metal figurines of the same type characteristic of Macedonian grave-goods, as are sanctuary of Masjid-i Sulaiman. subject discovered among the votive offerings at the well go back directly to the time of This direct link with Macedonian culture may Central Asiatic reworking of the Alexander, but in the Nisa harpy we see a genuine

W-i

Figure

5.

Old Nisa, reconstruction of a parade

shield.

Figure 6.

Old Nisa,

silver harpy.

168

Antonio Invemizzi

THE CULTURE OF PARTHIAN NISA
beings, such as to grant

169

called together their seats for some particular celebration, the architectural uniformity leads us to envisage a compact social order, a more or less egalitarian relationship among the elite, who, we may imagine to have formed the social hierarchy of the peoples following the Arsacids.

tamed close indeed took

been a venue for the holding of ceremonies that, in the new imperial context main ties with the past for the Arsacid aristocracy. If those

Turning to the classification of the architectural layouts, the Square House is related
to earlier Central Asiatic Iranian traditions, while the Tower Temple apparently has no particular parallels, and the Round Hall also presents problems. The cultural presup positions behind the Round Hall may, in fact, be sought in two different directions since, if its circular configuration immediately recalls a burial mound, the relationship between the main elements of a funeral mound, whether Scythian

him the right to be introduced into the divine court in the Goddess herself. From this or from some similar starting point, presence of the Great once they had acquired imperial dignity in the Asian political world, the Arsacids, could easily have developed the complex conceptions and forms celebrating their regal function, such as appear at Old Nisa. That the Arsacid aristocracy had not forgotten the themes of their ancestral culture, on the other hand, is documented by numerous artefacts. Among the silver Parthian treasures in the Getty Museum in Malibu there are representations of animal contest
scenes (Fig. 3), which clearly reveal that their origins lie in the animal art of the steppes and ancient Near East, in their composition as well as in their meaning; and it is no less
significant that these

themes should be chosen to embellish important parts of the har-

or

Macedonian—

room within— is reversed at Nisa with a circular, practicable space within a square. The funeral architecture of the area bordering on the European steppes calls to mind the monumental circular frescoed room of
the burial

circular without, but with a square, sealed

ness of Parthian horses, whose vital role in society the aristocracy introduced into the new state from their traditional way of life on the steppes.

mound

at

Kazanlak,

in

Thrace. But on account of both

its

shape as well

as

its

at Nisa appears to be the tholos of the palace at Vergina, the only circular space in a succession of quadrangular rooms and which perhaps served as a palatine

ritual

and non-funerary

which was decorated with figures, and which managed to escape the early looters of the Square House, is a disc bearing on one side a stag's head in relief in the typical position assumed in the art

And this is not all:

at

Nisa itself, one of the few metal

artefacts

functions, the closest parallel to the

room

chapel.

The
express

of the steppes by a stag downed by a predator: head lowered, one leg stretched forward, the other bent under the body (Fig. 4). The execution of the Nisa stag, however, has no exact parallel in the art of the steppes and must be interpreted as a reworking of those
characteristics in the light

size

and

variety of the buildings, not

easily understandable in
its

a

state

tombs but memorials, relate to displays and imperial society which had perfected a language to
high rank
rul-

of an aesthetic

sensitivity consistent

with the Parthian age.

We

cannot, however, speculate further about what relationship this singular image

particular concept of kingship in the light of a situation in which the sovereign now had to conduct himself, faced with other sovereigns of equally

whether this typically impeprobably born during the expansion of Arsacid Parthia, was coloured with a certain tone, a certain nuance the Arsacids brought with them from the by-nowdistant world of the steppes, or at least whether within this latter culture there were already the premises which were able to facilitate its grandiose future development. The Scythian art of Eurasia is dominated by the privileged relationship of man with
rial exigency,

ing in various degrees of relationship with the legacy of the former and the disintegrating Seleucid empire. But we may wonder

Achaemenid empire

the part of a Scythian, a prince or a chief, to a goddess, perhaps this people's greatest divinity, as

fraternizing, and from Greek mythology. In this repertoire of scenes showing human activity we also find, often in the form of little gold plaques to be sewn on clothes, a male figure, sometimes with a cup in hand, standing in front of a female figure on a throne. Clearly this is an act of homage or worship on

animals, which in fact feature as protagonists in many works. But on the European steppes, partly as a result of the close relationship with the Greek colonies on the Black Sea coast, man is often centre stage. In these depictions man does not interact only with animals but also with other men, with whom he is shown fighting or
there are also depictions of scenes

mentioned by Herodotus. It can be deduced therefore that this male figure, the chief, the prince, was recognized as having a special status above that of ordinary human

Figure 3.

Parthian phalera, Getty

Museum,

Figure 4.

Old Nisa,

disc with a stag protome.

Malibu.

1'

Antonio Invemizzi

THE CULTURE OF PARTHIAN NISA

167

Arsacid

and must therefore be strictly and exclusively related to a royal have been found in various parts of the monumental complex in the southern part of Old Nisa, and provide the most direct and significant concrete proof that religious ceremonies took place in these buildings, even though their use and their
parallels elsewhere,
cult,

within their Unlike Assyrian sovereigns, who were buried in modest hypogea whose grandiose funeral temples were quite palace walls, and unlike the pharaohs, Kings, the Achaemenid separate from their tombs hidden away in the Valley of the

meaning are by no means clear. So Old Nisa was not the Central Asiatic seat of the Parthian kings, but their sanctuary, the monumental mausoleum in which the royal cult was celebrated, whether the tombs of the sovereigns were there or not. The walls protecting this mausoleum, with
specific

mountains, and it is their own kings reinforced earlier traditions from the Iranian for the tombs that provide a concrete monument to their memory and a precise focus the vicinity of the capital. The great burial cult, being situated in precise places in

mounds of
ilar

the European steppes,
visibility in

on

the other hand,

show

that there

was

also a sim-

by the configuration of the sloping terrain, give the complex had no more of a military purpose than the walls protecting the picturesque series of terraces of the dynastic temple of Kanishka I at Surkh Kotal in Bactria.
the appearance of a citadel, but they

their massive bulk set off

Scythian society in the classical age. As for the West, the the royal palace of royal Macedonian burial mounds are situated not far from was brought back by Ptolemy I to his city, where Vergina, while the body of Alexander was to become an object of veneration. The ways in which the Ptolemaic

concern with

his

tomb

Now, the establishment
define
its

rituals

in itself of the principle of a royal cult and the concern to imply conceptions that attribute a special status to the sovereign and

are typical of Oriental empires. The demand belongs, in various ways, both to the Achaemenid and Ancient Oriental traditions as well as to the recent Macedonian and Seleucid traditions, the former influenced in certain ways. The royal celebrations naturally took different forms, but the need to enlarge upon them with powerfully visual architectural structures was understandably frequently and greatly sought after; it is found in various places and in specific forms. It may seem inappropriate to compare Old Nisa with other evidence, as neither the site nor the form of the royal burials are known, but even a superficial list may provide us with useful pointers.

long-established pracdynasty expressed their kingship later followed to some extent Seleucids, rulers of a composite tices in the country. In Asia, however, it was the affirming their legitimacy, the need to estabempire, who particularly felt, as a way of blessed with great cohesive political lish a real and genuine royal cult potentially
in which to counteract any possible centrifugal movements. The forms in Asia, with regard the royal Seleucid cult manifested itself seem, therefore, quite new as their displacement from the actual places to both the number of ritual sites as well of the Seleucids took place in temples scattered throughout the

power and

fit

of burial. The cult indicates that empire, like the temples of the gods, and the inscription at Nihavand at Laodicea. one stood on the Iranian plateau, There is no direct evidence of the Seleucid cult in Mesopotamia or in Iran but, at Nisa once about the middle of the second century BC the Arsacid sanctuary complex precise location nor again appears as something new. Even though we know neither the
the form of these royal tombs,
it is

above

all

the mighty encircling walls which

imme-

surrounding area, as prodiately strike us as clearly demarcating the complex from the protecting both the viding a unifying element for what lies within and, of course, as purpose of the wealth of the spiritual conceptions that find expression in the intended material wealth of the precious objects in use or constructions within, as well as the consequently, deposited. The totally new size and variety of the types of buildings and,

system and of their respective rites, imply a mature detailed elaboration of a conceptual suited to the great a specifically Arsacid ceremonial expression of kingship in forms

new empire.

We need not concern ourselves here with commenting in detail on the specific funcappear very diftions or the cultural relationships between the internal buildings, which interpretation that are by no ferent from one another and which pose problems of of the monu-

means
ritual

resolved.

The layout of each of the buildings

in the southern part

and mental complex seems, in fact, to offer a different response to specific ceremonial interpreted while the Square House in the northern part, which I have
demands,
Figure 2.

Old Nisa,

votive

gypsum

balls.

particular for banas a building constructed originally specifically for meetings, and in forms dear to the Iranian tradition, have quets and symposia, could, in architectural

164

Antonio Invernizzi

THE CULTURE OF PARTHIAN NISA

165

the relationship of the Arsacids to their
results

new country. And from
vital

of the excavations at Nisa are of

this point of view the importance, providing a corpus of widely

useful data.

The very existence of a monumental complex such as that at Old Nisa (Fig. 1), built by royal command, is indicative of a profound process of change in the new ruling class. Although the complex is known only in incomplete form, it invites us to wonder what might have been its possible prototypes, the specific architectural and artistic
influences, the spiritual links with the past

and the relationship with pre-Arsacid

Parthia.

Numerous

questions arise,

all

central to the process of interpretation, ques-

tions which are not easily answered, but asking

them may

at least help to formulate

concrete lines of research.

The first aspect which may provide information is the actual foundation of Old Nisa and its location. It appears as a fortified complex near the city of New Nisa, which is itself surrounded by massive walls. But it does not appear to be an acropolis

in either the

Greek manner or
is

in that of ancient
it

kingdoms of the East.

Its

con-

nection with the city

not as close as

may have been

at Persepolis, if the inhabited

areas of the latter extended as far as the foot of the royal
certainly
walls,

Achaemenid terrace, and it was for Assyrian palaces, which were situated within the city a solution which was also adopted in various complexes in ancient Central
was not
as
it

Asia.

Old Nisa, on one Kopet Dagh, was very probably determined by the desire to situate it on an elevated location on the plain beside which the city of Nisa (New Nisa) had been flourishing for some time. But it is no
of the
last

Of course

the precise choice of the site for the construction of
hills at

spurs in the belt of low

the foot of

less true that there is
t

nothing to lead us to think of the creation of a single

fortified

complex connecting the towered walls of Old Nisa to the massive walls of the city. The little remaining evidence provided by ancient remains around what is today the village of Bagir, does not shed light on the character of any possible occupation of
this

land in the Parthian era.
results of the most recent excavations increasingly support the hypothecomplex of Old Nisa probably functioned as a sanctuary for Arsacid

Now, the
sis

that the

kings.

The discovery

in the

Round Room of a fragmentary
I,

clay

head

(Fig. 10), that

has recently allowed the present writer to suggest that this room was really the mausoleum of the great king who gave the name of Mithradatkirt to the complex, and that it was a monument to the founder of both the empire and of Old Nisa. The tombs of the kings mentioned by
Isidorus of
far

may be

interpreted as a portrait of Mithradates

Charax were not

necessarily situated within these walls

no collaborating archaeological evidence has been found but it is likely that the monumental buildings so far excavated sprang from the desire to honour the memory of the great Arsacid kings by means of actual ritual ceremonies. These were quite unique in character. In them were found little gypsum balls with their surfaces often bearing impressions of coins (Fig. 2). These unusual finds, which seem to have no

or, at least,

so
100
50

m

m

figure 1.

Old Nisa, general plan.

162

Guy Lecuyot

CI. 1990. 'Greeks in Afghanistan: A'i Khanum'. In J. -P. Descceudres (ed.), Greek Colonists and Native Populations, Proceedings of the First Australian Congress of Classical Archaeology held in honour of A. D. Trendall, Sydney 9-14 July 1985, Canberra, 329-42, pis 36-9. SCHLUMBERGER, D. 1965. Ai Khanoum, une ville hellenistique en Afghanistan', CRAI, 36-46 SCHLUMBERGER, D. & BERNARD, P. 1965. Ai Khanoum', BCH 85, 590-657. VEUVE, S. 1987. Le Gymndse, MDAFA XXX, Fouffles a"Ai Khanoum VI, Paris.

RAPIN,

University of Turin

the Eurasian Classical sources tell us that the Arsacids arrived in Parthia from were the most eminent family. They steppes, together with the peoples of which they Seleucids, and founded a little state seized power in the province, at that time under the Their political which Mithradates I and his successors transformed into a great empire.

more commonly known to takeover in Parthia was so successful that the Arsacids are Parthians, even after their expansion westus, as they already were to the ancients, as the royal palace of wards made the provinces of Mesopotamia and Babylonia and with Rome. Indeed, it is with these to the story of their relationship
Ctesiphon central
events in the

our view West that ancient historians mainly concerned themselves. But first-hand caused by a lack of of Parthian history is vitiated by the obvious imbalance because, unfortunately, what information we have is information from the East
extremely fragmentary.

we can are helped by the archaeological evidence, thanks to which corresponParthia had ask ourselves whether this political takeover of the Arsacids in what these might have been. The same historical cultural parallels, and if so,
Fortunately,

we

ding

be based because, however sources provide evidence on which a tentative reply may major stages of the early history flimsy they may be, they nevertheless record the two
of the Arsacids, the stages of a history which
is

not only

political but also socio-cultural.

which must have been substantially In fact, the change from life on not matter whether nomadic, to a settled way of life which is that of a state— it does to the plains and deserts based on Achaemenid or Hellenistic models, or whether open
the steppes,

of Central

have had a profound effect on the social life (forme delta must also have involved fundamental vita) of the new arrivals. This new way of life part at least, had become aspects of thought and culture in this population which, in

Asia— must

the

new ruling class in Parthia. profit if The same available written documents may be consulted with much greater

archaeological research. In fact, it is placed in the relevant cultural context provided by fundamental to any attempt to discover above all the archaeological evidence which is
The "Rritish AcaHpmv

j„

1** \K1

1

11

if=i

7007.

242

Michael Airam
K. 1990. 'Sapur Kdnig von Iran - Faktum oder Irrtum', Schweizerische

MOSIG-WALBURG,

I.

Ein Abschied von Ardamitra', Numismatische Zeitschrift 102, 167-70. NIK1TIN, A. 1994b, 'Coins of the last Indo-Parthian king of Sakastan (A Farewell to Ardamitra)',
South Asian Studies
10, 67-9.

Namismatische Rundschau 69, 103-26, pi. 17. NIK1TIN, A. 1994a. 'Die Munzen des letzten indo-parthischen Konigs von STstan

NIKITIN, A.

1999. 'Notes

D. E. Klimburg-S alter

(eds), Coins,

on the Chronology of the Kushano-Sasanian Kingdom'. In M. Alram & Art and Chronology Essays on the pre-Islamic History of the

Tndo-Iranian Borderlands, Vienna, 259-63.

NOLDEKE, Th.
PANAINO,
201],

1879. Geschichte der Perser
iibersetzt

undAraber zur Zeit. der Sasaniden

—Aus der arabischen
*1

Chronik des Tabari

und mil

ausfiihrlichen Erlauterungen versehen, Leiden.

A. 2004. Astral Characters of Kingship in the Sasanian and Byzantine Worlds', Convegno internazionale La Persia e Bisanzio (Roma, 14-18 ottobre2002) [Atti dei convegni Lincei

Roma, 555-94.

'CNRS, Paris

PARUCK, F. D. J. 1924. Sasanian Coins, Bombay [reprinted New Dehli 1976]. RAPSON, E. J. 1904. Ancient Silver Coins from Baluchistan', Numismatic Chronicle Series 4, vol. 4,
311-25.

SACHAU,

E. 1915. Die Chronik von Arbela. Ein Beitrag zur Kenntnis des altesten Christentums im
6], Berlin.

Orient [Abhandhmgen der Preufiischen Akademie der Wissenschaften,

Jonathan Lee

SCHINDEL,
Vienna.

N. 2004. Sylloge

Nummorum

Sasanidarum,

III/l, III/2,

Shapur Tl-Kawad 1, 2 volumes,

SCHIPPMANN,

K. 1990. Grundziige der Geschichte des sasanidischen Reiches, Darmstadt. SENIOR, R. C. 2001. Indo-Scythian Coins and History, 3 volumes, London, SUNDERMANN, W. 1990. 'Shapur's coronation: The evidence of the Cologne Mani Codex
reconsidered

Afghanistan remains at the frontier of Central Asian archaeology, a country which still throws up remarkable discoveries, forcing scholars to rethink prevailing
historical presumptions. Despite twenty-five years

of

civil

war, discoveries such as the

VON GALL, H.
WIDENGREN,
WINTER,

and compared with other texts', Bulletin of the Asia Institute 4, 295-9. 1990. Das Reiterkatnpfbild in der iranischen und iranisch beeinflufiten Kunst parthischer
G. 1971. 'The estabishment of the Sasanian dynasty in the light of

Bactrian inscriptions of Tang-i Safedak and Rabatak, as well as other Bactrian and early Buddhist manuscripts, continue to appear. Remarkably, most of these discoveries
are not the result of systematic archaeological surveys but are usually casual finds

und. sasanidischer Zeit [Teheraner Forschimgen, VI], Berlin,

by

new

evidence',

Atti del convegno internazionale sul tenia:

La

Persia nel medioevo,

Roma, 31 marzo-5

villagers.

aprile 1970,

Rome, 711-82.
E.

& DIGNAS,

B.

2001.
Berlin.

Rom

und des

Perserreich

Zwei Weltmdchte zwischen

^Confrontation

und Koexistenz,

Rag-i Bibi, northern Afghanistan, which is the subject of this paper, is a further example of such 'casual' finds (PI. 6). Indeed, were it not for a series first intiof coincidences the carving would still not be known to foreign scholars.

The rock carving

at

My

mation of the existence of this carving

came
time
I

Museum held an Afghanistan day. At the

December 2002 when the British had recently returned from an expediin

tion to Yakaulang, Central Afghanistan, to recover the Bactrian inscription of Tang-i Safedak. The conveners of the conference kindly allowed me a ten minute 'slot' at the

end of the morning's proceedings to report on the discovery. An Afghan journalist, Najibullah Razaq, who was in the audience, later approached me and showed me video footage of the site which he and a BBC correspondent had visited earlier in 2002. Mr Razaq later forwarded additional photographs which showed a horse with rider at
the gallop
clothing,

and a number of standing or mounted attendants. From the it appeared that the carving was Kushan or Kushano-Sasanian.

attendants'

1

Francois Ory is a draughtsman and

Dr Philippe Martinez is an archaeologist and computer engineer, both at the

CNRS (UMR 8546).
rfthe Rritivh ArnJemv

13V 74V7.67.

© The British Academy 2007.

240

Michael A Iram

ARDASHIR'S EASTERN CAMPAIGN
ALTHEIM-STIEHL,
R. 1978. 'Das
friiheste

241

Ardashir the 'king of kings' after he had removed Farn-Sasan from Sakastan, probably
in the 230s.

Datum

der sasanidischen Geschichte, vermittelt durch

Archaologische Mitteilungen die Zeitangabe der mittelpersisch-parthischen Inschrift aus BTsapur',

From the area of Sakastan and Arachosia there
of Shapur
I

are also a

number of

stray finds of

Farn-Sasan's tetradrachms, Ardashir's throne-successor coppers and large copper coins

which might have circulated together in that area reported by Rapson (1904), & Ibrahim (1978) from the Kandahar Museum. Moreover, Mitchiner (1969) published part of a small hoard containing copper tetradrachms of the last Indo-Parthian kings, Pakores and Farn-Sasan, as well as large coppers of Shapur I probably overstruck on Ardashir's 'throne-successor type'. However, these

aus Iran 11, 113-16. BACK, M. 1978. Die sassanidischen Staatsinschriften. Studien zur Orthographie und Phonologic des mittelpersischen Mittelpersischen der Inschriften zusammen mit einem etymologischen Index des Inschriften (Acta Iranica 1 8), Leiden. Wortgutes und einem Textcorpus der behandelten

and by MacDowall

CHAUMONT,
(eds),

M.-L. 1974. 'Coregence et avenement de Shahpuhr I'. In P. Gignoux & A. Tafazzoli Memorial Jean de Menasce, Leuven, 133-46. CHAUMONT, M.-L. 1979. 'A propos de la chute de Hatra et du couronnement de Shapur ler', Acta Antiqua Academiae Scientiarum Hangaricae 27, 207-37.

FELIX, W.
Wien.

1985. Antike Literarische Quellen zur Aufienpolitik des Sasdnidenstaates, I (224-309),

finds are far too scarce to
in Sakastan.

be used as firm evidence for localizing the mint of these issues

GHIRSHMAN,

R. 1975. 'Chapour

I,

"Roi des

rois" sans couronne',

Monumentum

H.S. Nyberg

1

Together with Merv, Sakastan became an important strategic point on the eastern borderlands of the Sasanian empire. This is underlined by the fact that in both provinces mints were established in earliest Sasanian times which issued imperial gold

[Acta Iranica, 4], Leiden-Teheran-Liege, 257-67, pi. XXII. GOBL, R. 1954. 'Aufbau der Miinzpragung'. In F. Altheim & R. Stiehl (eds), Ein Feudalismus unter den Sasaniden und ihren Nachbam, Wiesbaden, 51-128.

asiatischer

Staat—

and

silver coins.

From Merv signed dinars and unsigned silver drachms are known from
(Alram & Gyselen 2003, pi. 21, A9, A 10 and pi. 35, Nikitin 1999, nos 26 and 27). The earliest signed drachms from
I

GOBL, R. GOBL, R.

1971. Sasanian Numismatics, Braunschweig.

as early as the reign of Shapur

phii-hist. 1983. 'Die Titel der ersten beiden Sasaniden auf ihre Miinzen', Anzeiger der Akademie der Wissenschaften, Veroffentlichungen der Numismatischen Klasse der Osterrechischen

A51; Loginov

&

Kommission

14, 290-8, pis I-III.

Sakastan are those of Shapur's son, Bahram
90-6). It
is

I

(273-276) (Nikitin 1999,
II

fig. 1).

The mint

of Sakastan continued to issue coins up to Shapur
retained, for

(309-379)

(cf.

Schindel 2004, nos

interesting to note that the tradition of minting large

example by Hormizd

II

(303-309) (Gobi 1971,

pi. 5,

copper coins was also nos 86 and 87), whose

O. 1996. 'Une monnaie en or du souverain Indo-Parthe 9-3 1 Abdagases IF, Studia Iranica 25, 2 1 GRENET, F. BOPEARACHCHI, O. 1999. 'Une nouvelle monnaie en or d'Abdagases F, Studia

GRENET,

F.

& BOPEARACHCHI,

&

Iranica 28, 73-82.

HARMATTA,
HINZ, W.
1

J.

1965.

'Minor Bactrian Inscriptions

5 ,

Acta Antiqua Academiae Scientiarum

left, a phenomenon which should be regarded as a special characteristic of the Sakastan mint and not as an error made by the die engraver (Figs 23 and 24) (cf.

portrait faces

Hungaricae 13, 149-205.
969. Altiranische

Funde und Forschungen,

Berlin.
I.

Schindel 2004,

1,

74; -one

of these coins was also analysed, Table
all

1,

no. 18,

and shows

HUYSE,

P.

1999. Die dreisprachige Inschrift Sdhbuhrs

an der Ka'ba-i Zardust (SKZ), [Corpus

a similar composition to

the other copper coins originating in Sakastan).

Inscriptionum Iranicarum, III.— Pahlavi Texts, I], London. KAWERAU, P. 1985. Die Chronik von Arbela [Corpus Scriptorum Christianorum Orientalium 467-8],

Leuven.

KETTENHOFEN,
Inschrift

CMC
NPIIN

Codex Manichaicus Alram 1986.
Sylloge

Coloniensis.

E. 1982. Die romisch-persischen Kriege des 3. Jahrhunderts nach Chr. Nach der Sahpuhrs an der Ka'be-ye Zardost (SKZ), Wiesbaden. KETTENHOFEN, E. 1995. 'Die Eroberung von Nisibis und Karrhai durch die Sasaniden in der Zeit 159-77. Kaiser Maximins (235/236 n. Chr.)', Iranica Antigua 30 (Festschrift K. Schippmann, II), ROMER, C. 1988. Der Kolner Mani-Kodex—Uber das Werden seines Leibes, KOENEN, L. &
Kritische Edition, Opladen.

SNS SNS SNS III TAVO

Nummorum Sasanidarum.
Gyselen 2003.

Alram

&

LOGINOV,

S.

D.

&

NIKITIN, A.

B. 1993. 'Sasanian coins of the third century

from Merv',

Schindel 2004.

Mesopotamia

27, 225^11, figs 1-5.

Tubinger Atlas des Vorderen Orients.

LUKONIN, V. LUKONIN, V.
Materialgnindlagen zu den iranischen

ALRAM, M,

1986.

Nomina Propria

Iranica in Nummis.

106-17. G. 1968. 'Monnaie d'Ardachir I et fart officiel sasanide', Iranica Antiqua 8, G. 1969. Kultura Sasanidskogo Irana, Moscow. MACDOWALL, D. W. & IBRAHIM, M. 1978. 'Pre-Islamic coins in Kandahar Museum', Afghan
Studies 1,61-71.

Personennamen auf antiken Miinzen Personennamenbuch, IV), Vienna.

(M.

Mayrhofer

&

R.

Schmitt

(eds),

Iranisches

MITCHINER, M.
301^1.

1969. 'A small hoard containing coins of Arda-Mitra', Numismatic Chronicle 9,

ALRAM, M. & GYSELEN,
(Ardashir I-Shapur
I),

R. 2003. Sylloge

Nummorum

Sasanidarum, Paris-Berlin-Wien,

1

Vienna.

MOSIG-WALBURG,

K. 1980. 'Bisher nicht beachtete Miinzen Sapurs
3, 1

I',

Boreas— Miinstersche

Beitrage zur Archaologie

17-26.

238
copper
coins'.

Michael Airam

ARDASHIR'S EASTERN CAMPAIGN
(cf.
r-l

239

R. Linke.

M.

Schreiner and J.-N. Barrandon in
J.

A first set of analyses was carried out for SNS I the contributions by Alram & Gyselen 2003, 70-90), and a
a
+->

t3

o o
C-l

s o

o
V

?
r-l

9
s 9

©, o"

o

O

tS

o
o"

o o
r-l
i

o
r-l

r-l

©

© o
9"

r-l

1

i

o ©n
o"

© O
9~

o ©

r-i

©
o"
1

r-l

© ©
o"

V

?
r-l

V

9
r-i

V

V

second was executed by
I

Riederer (Rathgen-Forschungslabor, Berlin) in spring 2004 using Atomic Absorption Spectrometry (Table 1). Riederer *s investigations (for which

am

<

3

extremely grateful) corroborate the results published in

SNS

©„ o" V

o

o
?"

(M

o
r-i

9 ©
o

V

?"

©_ o"

V

© O o

r-l

©
o"

rs

o

©
t

V

9
©
o"

© © © o 9
1

g o o

r-l

©
o"

r-l

©
J

V

© 9 ©
o"

©^
o"

o
r-l

o o
r-i

o
r-l

V

9
©. o"

©_
o"

9
©

I fully:

The most

K!

normal bronze coins is a relatively high percentage of lead ranging around the 20% mark, and this is also the case with the following issues of his successor Shapur I where the lead content is sometimes over 30% (cf. AJram & Gyselen 2003, 80-1 (Linke and Schreiner) and 89, fig. 1 (Barrandon)). The
characteristic
in Ardashir's

component

a
t/3 •±4

U

O

o

o

o
in IN

o

9
V,
r-l

9
in
c-l

©
V
(S

o
s

©
o"
in
r-l

o

V

9
>n
r-l

V

9
i^i r-l

i

V

9
©" V

u-i

in
r-l

S
<

©„ o"

©

o
o"

O
V

o
i

o
o"
1
i

V

9
o

?
©
o"

o"

r

V

9
©
•*
r-l

©^ o"

V

V

o^ o"

o

r-l

o^
o"
1

o
r-l

V

9
©

V

lead content of the 'throne-successor coins', however, normally

lies

nos 1-8) as well as of those copper coins of Shapur T, which all some of them clearly struck on Ardashir's 'throne-successor type' as already stated (Table 1, nos 13-17). A further criterion is the weight of the copper 'throne-successor coins': with a mean weight of 11.50 g (Shapur I's overstrikes have approximately the same weight) they are much lighter than all the other bronze issues of Ardashir 1, which I have attributed to 'Ctesiphon' and 'Ramadan' and which weigh around 17 g (cf. Alram & Gyselen 2003, 163 ((., tables 9-13).
Finally,

below 1% (Table 1 seem to be overstrikes

o
•s

©
V
in

©

©

in

ON

o
?"

©

o
V

?
>n

V

?
o
in

©
V
to

o
V
in

o o
in

©"

o ©"
5 o
o"

© V
in

o
r-l

©
9"
c?

©
OS

O

©
tfl

©

o
o"

o" V
co

9
SO o o

V

o V
o o
o"

o

©
V
Of

In

o
9"

o
©,
o"

©

©
o"

in

V

o
V
r-l

o o ©"
in

>n

©^

o
o"

m o

in

V

V

9
r-l

© V
P.

©

© o

in

©,

©"

©"

V

•o
SO

<

so

o
o"

©

©

©"

o
o"

o
o"

r-l

©,
o"

o

M3

to

©"

© © ©
in

o on © ©'
o"

n
©
r-l

©

o"

© ©
©__

to

SO

©
o'

©
©'

© ©_
r--

o"

Z
2

© V
o

©

©

o
o" o"
vo

o

O
©"

9
r-l

V

© ©

©

o

©"

©„

i

©
©
o'
r-

1 1

©
o"

©^

V
©^

©

j

9 © © © ^

metal analysis of four coins of Farn-Sasan (Table

I,

nos 9-12) has shown

that their chemical composition

§
a
a
CO

ft

©
vo

o"

o"
80 o o

©"
SO © o © © © ©
MS

j

©"
SO © ©_

©
j

r-l

1

o"
CO o o^

©"

©"

©"

j

is quite similar to those of the 'throne-successor coins': they also consist of nearly pure copper and a very low percentage of lead. The lighter weight of the 'throne-successor coins' also corresponds closely to the weight of Farn-

S3

SI

o o o o o, ©"

o

© o

1

o o

©
©"
©_

o"

o"
r-l

©
o"

j

;

o"

©"

© ©
©

© ©^
o"

<

d"

i

©
©
9"
>n

Sasan's second series with larger flans, which as mentioned above has a

mean weight of

J3

©"

dig.
To sum up, the differences in chemical composition between Ardashir's bronze and his 'throne-successor coins' might be the result of a different production process, and can therefore be taken as an indication that these issues were produced not at Ardashir's two main mints ('Ctesiphon' and 'Hamadan'), but at another mint which
coins
s
CO

9
<n
CI

V

o"

©„ ©"

r-i

oa

V

o"

8.

©
as

o ©^ ©
o o o

©"

Os r-

©
m
r-l

© 9
>n
IN

s d r3 s H
'd'
>-

"J.

"T.

r*

c3

o"

Z
^^
T)

in
r-l

•o
IN

©"

9

V

9
Os
so"

?"

o"

V

9 o

o"

V

o o o
o

in
r-l,

OS
r-i

in
r-i

in
r-1,

r-l

a-

o"

©
o-"

V

?
o.
r-l

9
CO

o"

9
o
Os"

o"

V
CO

9
OS CN
co"

o"

V

©^ o"

© V
°\
aOs

T3

©
s of
Os
os

oo OS

SO
r-;

o
co"

ao-"

OS
a-'

S
J5
•S
S-J

a-?

ON

o-.

ON

CO a-

OS

as

OS

as

Os

o\
Os

OS

d
xj

can probably be localized in the Sakastan area, where the coins of Farn-Sasan were also produced.

d

s
SO
C-l

Z
co

c-,

to

CO IN

ci
r-l
r-i

>
r-l

_>

1

SO

z
3
'3

W
ob ob on
r-l

t~-

1

5

"^ in

§

U-,

5^
fell

T3

A further

argument should be noted.

It

was Skjsrvg who

first

realized that the
bi)

50

ob

inscriptions of the 'throne-successor coins' contain a strange mixture of Sasanian-

oh
C--

oil

oo
t--

and

C>'

t
©"

ob Os

w
r-l

ob so
r-l"

ob ON

Oi 73 -

'n

in

©"
so"

o"

o"

Parthian-type letters which in

some parts

— 5
>->

j:

display close similarities to the coin inscrip-

tions of Farn-Sasan, the last Indo-Parthian king of Sakastan (Skjsnwr in

Alram

&
o
S « ™

Tl
<M

Gyselen 2003, 55-6). Lukonin (1969, 40) was the first to assign these coins to Sakastan, even though his reading led him to wrongly attribute them to Ardashir, the king of Sakastan mentioned in the Shapur I's Ka'ba-I Zardusht inscription (SKZ).

^
id

§
a

5^

§

^

1

($

1 <
rn
j

g g g & <
<!
it-

55

^ ^

^
id

^
-a

^^
1

s
at
B)
CO

ea

8
CJ
i

J

ft

The basis of investigation is of course far too narrow to supply any definite answers,
but
it

Z> s CO
f-i

>

| >

s
>

s > >

CO

J

1

J

>

>
el

CB

Hi

V3

a

S3

does provide

initial

indications that Ardashir's 'throne-successor coins' probably
t.

B

e
M S
CO

originated

from the Sakastan

area. If this

is

true,

then the young prince on the obverse,

«
!« •a

a a a 3
a M
CO

.G

J3

who

displays such strong similarities to the beardless

might probably be identified

young man on Ardashir's reliefs, as Ardashir. the king of Sakastan, who was put in charge by

<
"*

B 1.

s

o
•<

<

E

a

<

a u «
(a,

s
M
fe.

s «
ft

u a
a.

«

S «
to

a
a. CO

u

3
ea

1
B o

CO

to

't

SO

K «

it
H
Pi

236
ist

Michael Alram

ARDASHIR'S EASTERN CAMPAIGN
.

237

Legende mit den derzeit zur Verfugung stehenden Materialien nicht lesbar. .'), although he was still certain that it was Shapur as crown prince who was depicted. While preparing my Iranisches Personennamenbuch (Alram 1986), nearly 20 years ago, I myself tried and failed to read this legend. Finally, during our work on the first volume of the Sylloge Nummorum Sasanidarum 1 tried once again, only to fail yet again. Even Skjaerv0, who prepared the palaeographical part of our publication, was unable to read it (Alram & Gyselen 2003, 55-6, Skj<srv0 thought the name of Ardashir' could perhaps be read on two specimens), as also were Carlo Cereti (Rome) and Dieter Weber
die

I

am unable to provide a solution to this problem here. My aim has been merely to

show

that the identification of the 'crown prince' Shapur, on Ardashir's so-cailed 'throne-successor coins', is not as certain as is always stated in numismatic literature. In

my

view there are rather strong similarities between the beardless young
reliefs,

man on

the

whose identity, 'throne-successor coins' and the beardless man on Ardashir's however, is somewhat unclear. I would like to point out that the famous high priest
Kartir
is

also always depicted as beardless;

he too

the beardless young

man on Ardashir's reliefs,
like

is wearing a similar tiara to that of decorated with a special nisan, but also

(Gottingen). Naturally this does not
it

in the future.

On the

other

mean that somebody might not be able to decipher hand we should be careful to guard against constructing
exist.

with a diadem band
Therefore the young
successor' coins

the tiara of the young

man on

the 'throne-successor' coins.

man on

must be aware of the fact that Sasanian coins sometimes bear totally blundered legends, as can be seen, for example, on the famous gold dinar of Shapur I from the mint of Merv (Alram & Gyselen 2003, pi. 35, A51), and
even in the reverse legends of the 'throne-successor coins' themselves Ardashir's appears in a corrupt form on most of the coins.
Besides the epigraphical problem, the question remains as to

things which definitely

do not

We

may

Ardashir's reliefs as well as the person on the 'throneregarded as a kind of priest (T am grateful to Shahrokh be

Razmjou of Tehran for making this helpful suggestion). The epigraphic and literary sources indicate that Shapur received

the diadem of the

name

place between April great king during the lifetime of his father in 240. This probably took according to the Bishapur inscription (SVS): the lighting of the and September, because

why Shapur should be

Shapur

depicted as a beardless young

man in

the 230s, only five or six years before he himself

became the 'king of kings'. In 223/224 he had already fought at his father's side in the battle of Hormizdagan, as recorded in literary sources as well as on the famous rock relief at Firuzabad, where he is depicted as a bearded man (cf. von Gall 1990, fig. 3, pi. 6). Shapur is also shown as a bearded man on the second relief of Firuzabad and in
Naqsh-i Rajab. According to the numismatic evidence
all
is

18, dated to 239/240 and according to the Mani Codex in Cologne, Seleucid Era = 1-10, it is stated that, when Mani was 24 years old, (i.e. year 551 of the diadem 12 April 240-1 April 241), Ardashir conquered Hatra and Shapur received the
fire is

CMC

(Koenen

& Romer

1988, 10;

cf.

Sundermann

1990, 295

and Huyse 1999,

2, 7;

Alram

&

these reliefs should be dated
to say to the last ten years of

same period as the 'throne-successor coins Ardashir's reign, between 230 and 240.
to the

5
,

that

Gyselen 2003, 149-52 also gives a full discussion of this ruled In this connection it has been repeatedly suggested that Ardashir and Shapur cf. also jointly (Chaumont 1974, for example, but later rejected Chaumont 1979; Kettenhofen 1982 and Sundermann 1990). Mosig-Walburg 1980 and 1990, as well as
literary In numismatic terms, the theory of a co-regency was created by linking the have sources to the so-called 'throne-successor coins', the problematic nature of which I

in a numismatic context).

is the beardless young man who normally stands behind the 'king of kings' holding a fan over Ardashir's head, as in the investiture reliefs of Firuzabad, Naqsh-i Rajab and Naqsh-i Rustam. In the battle scene of Firuzabad he is depicted on horseback fighting together with Ardashir and Shapur

A very prominent person in Ardashir's reliefs

already pointed out here. In

my opinion, they may be ruled out as proof of a joint reign
that he succeeded his father as sahan sah

by father and son. Shapur's coinage shows

against the Parthians. His characteristic insignia

is

a high tiara without a diadem band,

decorated with a special nisan. Hinz (1969) has identified this person as Ardashirs Knappe und Wedeltrager' which does not add significantly to our knowledge. In any
case the tiara
is

Eran ke cihr az yazdan without limitation. As Lukonin has pointed out, many of the copper issues of Ardashir's thronenos successor type were overstruck by Shapur I (Figs 19-22) (Alram & Gyselen 2003, why Shapur overstruck precisely those 5-A8; cf. Lukonin 1968, 108). Here we may ask
coins which were intended to bear his

rather similar to the headdress of the beardless

'throne-successor coins', here, however, the tiara

Gobi (1983, n. 11) was unable to identify any trace of it on the coin in question (Fig. 10) when I had the opportunity to examine it directly. Moreover, a beardless young man or boy is found

young man on the is decorated with a diadem band. believed he could detect a nisan on the diadem of the 'prince', but I

own image? Other

overstrikes are

known from

the mint of Merv, where

bronze coins of Shapur I were struck over issues Ardashir's issue of Merv mintage (coins of the local king were also struck over surmised that a large series of Ardashir's of Ardashir I). Loginov and Nikitin have coppers had been stored at the mint and was used as a ready stock of flans for striking

many

early

on the relief of Naqsh-i Rajab

together with a naked deity holding a club, identified by Walter Hinz (1969) as Yerethragna and Bahram, the eldest son of Shapur, who became 'king of kings' in 273. Finally, the type of a beardless young man is depicted on coin issues of Bahram II; here too he is usually identified as the crown prince (Gobi 1971,
typell/l;pl. 4, no. 54).

new coins on

the accession of Shapur I (Loginov

& Nikitin

1993, 229).

these lines In the case of the 'throne-successor coins', however, an explanation on shown that in their chemical composition does not seem to add up. Metal analysis has

other bronze the 'throne-successor coins' are quite different from all of Ardashir's described as 'pure coins, and because they are completely lacking in tin they can be

234
(SKZ), Ardashir installed a king
probably a
in

Michael Alram Sakastan

ARDASHIR'S EASTERN CAMPAIGN
also

235

who was

member of the royal

family (Huyse 1999, 54,

SKZ §41).

named Ardashir and was In Merv a sub-king

named Ardashir was also enthroned (at the beginning of his career Ardashir had already installed his brother Ardashir as king of Kerman). Ardashir also succeeded in
conquering al-Bahrain. It is not possible, however, to determine exactly when the Arabian campaign took place, and we are more or less forced to speculate. According to Tabari the conquest of al-Bahrain followed the eastern campaign, which in turn
followed the conquest of Ctesiphon; by contrast, the Nihayat places the Arabian campaign after the conquest of Ctesiphon (Widengren 1971, 752 ff.; Schippmann 1990,
17 ff., n. 33 with additional literature). Kettenhofen (TAVO, B V 1 1) dates the Arabian campaign to around 235, Harmatta (1965, 193) around 237/238. The history of Sakastan in the late Parthian and early Sasanian period is somewhat unclear. From the numismatic sources we know that a certain Farn-Sasan (Figs 13-18) ruled in this region, who might be regarded as contemporary with Ardashir I. On the obverse of his copper tetradrachms the bust of the king facing left is depicted in the

however, the authenticity of this coin seems to me highly questionable (Grenet Bopearachchi 1996 and 1999; Senior 2001, 2, 190). Farn-Sasan eventually reorganized the coinage of his predecessors in the Sakastan
coin type was introduced showing an interesting mixture of old IndoParthian and new Sasanian elements. For the coin inscriptions Farn-Sasan used the Parthian script, which is also found on the silver drachms of his great-grandfather
area.

&

A

new

Sanabares from Sakastan. As Nikitin (1994a) has shown, the coinage of Farn-Sasan can be divided into two main series. The first is struck on smaller flans (18-20 mm), with elaborately engraved images and legends and a mean weight of 7-8g (Figs 13-16). On some rare coins of this

The second series has larger flans (25 mm), often with blundered inscriptions, and a mean weight of c. 1 Ig (Fig. 18). It should
group the king's portrait faces right
(Fig. 17).

of the coin portraits of the Indo-Parthian kings. On well-preserved specimens we can see the left arm of the king holding an arrow in his hand (Fig. 16). As far as I know, this type has no parallel in the Indo-Parthian coinage but is, however, well known from
style

be noted, however, that the inscription on this second series, although blundered, might not be identical to that of the first series. Around 233 Indo-Parthian rule in Sakastan came to an end, and the region became an integral part of the Sasanian empire, placed under direct control of a king named Ardashir who was probably a member of the royal family or a close confidant of the 'king of kings'. As mentioned above, another subking named Ardashir was put in
charge in Merv, and a third king who was definitely a brother of Ardashir I, as reported by Tabari, and whose name was also Ardashir ruled over Kerman province (as attested

the coins of the
clearly copied

Kushan king Soter Megas.

On the reverse we see the Sasanian fire altar
for the
first

The legend is written

from coins of Ardashir I. in Parthian Middle Persian and was read correctly
it

time by Nikitin. Unusually,
altar 'Farn-Sasan, son of

obverse

at

1 1

left of the fire Adur-Sasan' (prnssn— BRY Hwrssn) and continues on the o'clock 'grandson of Tiridat, great-grandson of Sanabar, King of Kings'

starts

on

the reverse at 10 o'clock to the

Ka'ba-i Zardusht inscription §41, Huyse 1999, 54). These kings or provincial governors did not normally have the right to mint coins. This was the exclusive prerogative of the 'king of kings', and an important part of Ardashir's monetary reform in contrast to conditions in former Parthian times. There is
in the

Shapur

I's

(BRY BRY

tyrdty

BRY

npy s'nbry

MLKYN MLKA) (Nikitin
as well as the
title

1994a and

b).

only one exception to this rule

— apart from the Kushano-Sasanian coin
I

series

—when

The arrow

in the

hand of the king

'king of kings' indicate that

Farn-Sasan regarded himself not as vassal or friend of Ardashir but as his enemy, despite having chosen the Sasanian fire altar for the reverse of his coins.

the king of Merv mlwy MLKA, which might be

issued small bronze coins under Shapur

bearing the inscription

seen in the context of a long tradition of local bronze

The only person mentioned
is

in the coin inscription of

Farn-Sasan

whom we know

his great-grandfather Sanabares,

who

ruled the Indo-Parthian

half of the second century ad.

He

struck silver drachms in

kingdom in the first Sakastan and the so-called

& Nikitin 1993, nos 123 ff.). should return to the so-called 'throne-successor coins' (Figs 9-12) of were in any case minted in as I have attempted to show Ardashir I, which be dated roughly to between Ardashir's third minting phase and should therefore
coinage in that area (Loginov

Now we

Nike-tetradrachms in Arachosia (Alram 1986, nos 1191-6; Senior 2001, nos 261-5). The successors of Sanabares were Abdagases II and Pakores (Alram 1986, nos 1142; 1189-90 and Senior 2001, nos 233-5, 268-9). Both Icings issued silver drachms

and copper tetradrachms as well. After Pakores, however, the silver drachms disappeared and only the copper tetradrachms continued to be issued, but in an increasingly
barbarized style with totally blundered legends (Alram 1986, nos 1214-6 and Senior 2001, nos 21\-%). These coins seem to be the immediate forerunners of Farn-Sasan's tetradrachms. On some of these coins the king already holds an arrow (Senior 2001, no. 272. 3T); the same motif is also seen on a unique gold coin attributed to Abdagases II,

and stylistic peculiarities, it is not possible to determine whether they were struck in one of the two main mints or at another mint which we are not able to localize at present. Since Paruck (1924) the two busts on the obverse have been identified as Ardashir and his son and crown prince Shapur. Paruck also proposed a reading of the obverse legend which runs as follows, starting at 3 o'clock: 'Shahpuhri malka Airan minochitri'. With some variations and additions this reading was taken over by Gobi (1954), Lukonin (1968), Ghirshman (1975) and finally Mosig-Walburg (1980 and 1990). In
c.

229/230 and 239/240.

Due

to their typological

1983, however,

Gobi

stated that the reading of the inscription

is

doubtful ('Jedenfalls

232
This date

Michael Air am
is

ARDASHIR'S EASTERN CAMPAIGN
At a
later stage,

233

taken by Altheim-Stiehl as Year

1

of the reign of Ardashir and inserted

probably in 226/227, Group/Mint

C is added, which can be definitely

would then mean that the Shapur Fire was lit in 239/240, that the beginning of the unnamed era would fall in the year 205/206 and that the monument for Shapur I in Bishapur was erected in 262/263.
minting phase thus begins with his coronation as king of Fars in Stakhr and ends with his victory over Artabanus IV at the Battle of Hormizdagan in 223/224. It remains controversial, however, as to just when Ardashir's coronation as king of Stakhr took place—and thus when minting began in Fars (Phase 1). reliable terminus post quem is providedin any case by the unnamed 'Sasanian
first

into the equation given

by the Bishapur

inscription. This

identified with

Ctesiphon and which according to the typological evidence clearly

started later than the coinage in

Group/Mint
it

B.

From the

strategic point of

view

seems to

me very unlikely

that Ardashir started

Ardashir's

his eastern campaign before the conquest of Ctesiphon. Ctesiphon was the centre of power in the western part of the Arsacid Empire and therefore it must have been

Ardashir's primary aim to gain control of
this took nearly three years
is

Mesopotamia

as quickly as possible.

That

A

Era' mentioned

a sign that Arsacid resistance remained very strong even after the conquest of Media and the death of Artabanus IV. I thus prefer to follow the
course of events as outlined by Tabari.

in the Bishapur inscription (SVS) and which (according to Altheim-Stiehl 1978) began in the year 205/206. The absolute terminus ante quem is the Battle of Hormizdagan,

which
If

is

supposed

to

have taken place in 223/224.

one accepts Tabari's account, Ardashir proclaimed himself 'king of kings' immediately after the battle, while still on the field (Noldeke 1879, 15, according to the Nihayat, Ardashir adopted the title only after the conquest of Ctesiphon, cf. Widengren
is unclear. According to Tabari, from Media described a large arc through Adurbadagan / Atropatene, Nodshiragan / Adiabene to Asuristan / Assyria (Iraq), where he conquered

1971, 771).

The course of

events that then followed

Ardashir's subsequent advance

the capital of the Parthian Empire, Ctesiphon, in 226/227. order to prepare his great eastern campaign. This took

Abarshahr

at least as far as Merv.

However, the

Then he returned to Fars in him through Sakastan and account in the anonymous Nihayat al(cf.

Widengren (1971) accepts the version in the Nihayat and thus dates Ardashir's eastern campaign to before the conquest of Ctesiphon. I have only one numismatic argument against the version in the Nihayat; however, this does not prove that the account in the Nihayat is wrong, but can rather be taken as an indication that Tabari is right: this is the bronze coinage from the mint at Merv (Fig. 8), which starts only in Ardashir's third minting phase, thus in any case after, and not before, the conquest of Ctesiphon. If a causal relationship is assumed between the beginning of the coinage in Merv and Ardashir's eastern campaign, the numismatic evidence could that the eastern campaign took place after the indicate as reported by Tabari

By

contrast,

conquest of Ctesiphon.

Irabfi-ahbar al-Furs wa'l-'Ardb
after Ardashir's victory over

is different,

placing this eastern campaign immediately

Artabanus and the conquest of Media

Widengren

also accepts the version in the Nihayat and thus dates the eastern campaign to before the conquest of Ctesiphon; Harmatta 1965, 186 ff. thinks otherwise, dating the eastern campaign— in accordance with Tabari— as starting in 233, following the conflict with Severus Alexander; cf. also Kettenhofen 1995, 1

1971, 745 ff,

who

At

this

point

it is

of power as

known

65 ff. necessary to return once again to the dates of Ardashir's seizure from ancient records. The literary sources provide two different

The third and final period of Ardashir's rule, which may be roughly dated from c. ad 229/230 to c. 240 is marked primarily by conflict with Iran's traditional enemy, Rome, which started around 230. Evidently Ardashir felt he was now powerful enough to extend Sasanian expansion to the west as well. The first attack took place, as mentioned above, around 230: Nisibis was besieged, and Sasanian troops penetrated as far as Syria and Cappadocia (Felix 1985, 32 ff., discusses the sources; Winter & Dignas 2001, 87 ff. gives a summary of the events and additional references). The Roman emperor Severus Alexander launched a counter-offensive in the spring of 232. The
fighting probably
losses.

ended without a clear

victory,

with both sides suffering considerable

dates for the beginning of Ardashir's reign, which are perhaps related to two different events. One is the date of 223/224 given above (SVS; Tabari; Chronicle of Arbela, Acts of Syrian Martyrs), which is presumably related to his decisive victory over Artabanus at the Battle of Hormizdagan. On the other hand, Agathias (4, 24, 1 Keydell) and Elias of Nisibis (42, 16 ff. 91, 18 ff. Brooks) give year 538 of the Seleucid era (1 October 226-30 September 227) as the year the empire was founded. As early as 1879, Noldeke
;

Following Severus Alexander's withdrawal, there were three to four years of peace on the western front (Herodian 6, 6, 6) before a new advance into northern Mesopotamia took place in 235/236, in the course of which Nisibis and Carrhae were
captured by the Sasanians.

with the conquest of the imperial capital of Ctesiphon (Noldeke 1879, 409 ff., especially 411), a hypothesis that has been widely accepted by subsequent scholars and than which I have no better solution to offer. To a certain extent the numismatic evidence as outlined here corroborates the course of events. Ardashir's second minting phase, his first as 'king of kings', starts in 223/224 with coins of Group/Mint B which I have assigned to the mint of Hamadan.
this date

had connected

As Harmatta and Kettenhofen have

suggested, the eastern campaign of Ardashir

might have taken place within this time span, probably starting in 233 (cf. Harmatta 1965, 186 ff. and Kettenhofen 1995, 165 ff). According to Tabari, Ardashir started his campaign from Fars. He first marched eastwards to Sakastan and then to the north through Abarshahr at least as far as Merv, which was to hold a key strategic position in the north-east of the Sasanian empire (Kettenhofen 1995, 171, n. 74). According to
the
list

of Ardashir's dignitaries in the inscription of Shapur

I

at the Ka'ba-i Zardnsht

230

Michael Air am

ARDASHIR'S EASTERN CAMPAIGN
with Phase
3, the focus of

231
to

Type IT(3)/3a(2) coins may be divided into two groups (B and C) that apparently reflect two different mints. Group/Mint B (Fig. 3), which is in any case older than Group/Mint

minting

now

shifts

from Group/Mint B ('Hamadan')

Group/Mint C

('Ctesiphon'), a development that corresponds precisely with the course

C

have associated with the Median capital of Ecbatana/Hamadan, which had been the main mint for the drachm production of the Parthian Empire.

(Fig 4)

i

The

crucial point in establishing the location of

billon (base silver) tetradrachms (Fig. 5),

Group/Mint C is provided by the which are clearly a Parthian legacy and which

of political events. each that depict Ardashir with In both groups/mints there are two special issues perhaps related to certain religious (?) ceremonies (Alram special crowns and which are 2003, types TV/3a, V/3a (group/mint C) and types VI/3b and VII/3b (group/

& Gyselen
mint
B).
lished.

the Arsacids as a rule only minted in Seleucia-Ctesiphon. As I will explain later the capture of Ctesiphon presumably occurred in 226/227. It is likely that Ardashir 'also took over the mint in the same year, issuing mostly tetradrachms and small bronze coins in the Parthian tradition.

In Phase 3

it is

probable that the mint in provided by an

Merv (Group/Mint D) was

finally estab-

Evidence

is

edition of small bronze coins {Mil) of

Type

IIIa/3a,

whose

characteristic,

somewhat

coarser style distinguishes

them from the contempor-

PHASE 3, the final phase
new
[is]

probably produced as small change ary issues of the two main mints and which were based in this case solely on the evidence for local needs (Fig. 8). The location of Merv is
Nikitin 1993). of finds (Loginov Also part of Phase 3 are the so-called 'throne-successor coins' (Type VIII/3a). Drachms and large copper coins (JEll) of this type were issued (Figs 9-12). The main
focus of production

coin type

and 7). from the gods') has been added (the primary meaning of cihr is 'visible form, manifestation', according to Panaino 2004, 599). The typological criteria permit the conclusion that the bipartition of
(Figs 6

of Ardashir 's minting, is marked by the introduction of a on the obverse showing Ardashir with a covered, artificial hairstyle Also new is the legend to which kecihr azyazdan ('whose family/seed
visible

&

minting as

noted

in

Phase 2 continued and thus two mints were in operation. However, beginning

copper coins, of which there are three different series. Most of these issues were probably produced in the second half of the third minting Groups/Mints phase. The attribution of the 'throne-successor coins' to one of the two C and B remains open, although it cannot be ruled out that an additional mint may
is

clearly the

also have

been involved, perhaps located in the area of Sakastan, as we shall see below. The very end of Phase 3 can at present only be seen in Group/Mint C ('Ctesiphon').

I, drachm (3,77 g), type 1/1 (group/mint A, Stakhr). SNS I, I. 2, Ardashir I type IIa/2 (group/mint B, 'Hamadan'). SNS I, A2. 3. Ardashir I, drachm (4,03 g) type IIa/3a (group/mint B 'Ramadan'). SNS I, 8. 4. Ardashir T, drachm (4,17 g), type IIe/3a (group/mint C, 'Ctesiphon'). SNS I, 23. 5. Ardashir I, tetradrachm (13,70 g), type IIe/3a (group/mint C, 'Ctesiphon'). SNS I, 31. 6. Ardashir I, drachm (4,16 g), type TIIa/3a (group/mint C, 'Ctesiphon')!

Figures:

1.

Ardashir

The decisive typological and stylistic criterion for the classification of these latest issues which are delineated in small strokes, one is provided primarily by the altar flames,
above the other. These 'broken' flames are the only clearly definable element also found on the early coins of Shapur I and thus form the only typological bridge between Ardashir's and Shapur's coinage (Alram & Gyselen 2003, Ardashir I, phase 4: nos
246-58; Shapur
I,

dinar (8,47

g),

s

SNS
8.

I,

125. 7. Ardashir
I,

I,

drachm

(4,27 g), type IIIb/3b (group/mint B, 'Hamadan').

SNS

I,

208

style

A: nos A2, A6, All, 9-13).
in

Ardashir

AE/2

VITI/3a (Sakastan?).
1983, pi.
1, fig. 2.

SNS I, 259. 9. Ardashir I, drachm (3,60 g)' type A54. 10. Ardashir I, drachm (4,07 g), type VIIT/3a (Sakastan?). Gobi 11. Ardashir I, AE/1 (?), type VIII/3a (Sakastan?). SNS I, A56. 12. Ardashir I,
(3,17 g), type IIIa/3a (Merv).

SNS

I,

AE/1

(11,85 g), type VIII/3a (Sakastan?). SNS I, 239 (analysis cf. Table 1, no. 3). 13. Farn-Sasan^ copper tetradrachm (5,95 g), Sakastan. Miinzkabinett, Wien (analysis cf. Table 1, no. 9). 14. FarnSasan, copper tetradrachm (7,33 g), Sakastan. Miinzkabinett, Wien (analysis cf. Table 1, no.
10).

a chronological framework of absolute dates This inscription is is the famous inscription of Bishapur (SVS) (cf. Back 1978, 378-83). dated to the year 58 of an unnamed era; it is equated with Year 40 of the Ardashir Fire

Our main source for placing the coins

and Year 24 of the Shapur

Fire.

The

calculation

made by

Altheim-Stiehl (1978)

copper tetradrachm (6,94 g), Sakastan. SNS I, pi. 39, E19 (analysis cf. Table 1, no. 12). 16. Farn-Sasan, copper tetradrachm (?), Sakastan. American Numismatic Society, New York. 17. Farn-Sasan, copper tetradrachm (5,40 g), Sakastan. SNS I, pi. 39, E20. 18. Farn-Sasan, copper tetradrachm (11,04 g), Sakastan. SNS I, pi. 39, El 8 (analysis cf. Table 1, no. 1 1). 19. Shapur I, AE/1 (1 1 g), type Hal/la (style Abis; overstruck on Ardashir I, type VIII/3a). SNS I, A7. 20. Shapur I, AE/1 (13,89 g), type Hal/la (style Abis; overstrike). Miinzkabinett, Wien (analysis cf. Table 1, no. 1
5).

15. Farn-Sasan,

is the very date that assumes that the Ardashir Fire was lit Tabari named as the year in which Ardashir killed the Arsacid great king Artabanus IV in the battle of Hormizdagan and proclaimed himself 'king of kings' of Iran (Noldeke

in the year 223/224. This

AE/1 (10,52 g), type Hal/la (style Abis; overstrike). Miinzkabinett, Wien (analysis cf. Table 1, no. 16). 22. Shapur I, AE/1 (12,16 g), type Hal/la (style Abis; overstrike). SNS I, 5 (analysis cf. Table 1, no. 17). 23. Hormizd II, AE/1 (9,29 g), Sakastan. Miinzkabinett, Wien (analysis cf. Table 1, no. 18). 24. Shapur II, AE/1 (6,41 g), type Ill/lb (Sakastan; cf. SNS III, A 15). Miinzkabinett, Wien.
21.
I,

Shapur

transmitted by the Chronicle of Arbela, which records that the reign of the Parthians ended in the year 535 of the Seleucid era (223/224) (Sachau 1915, 16; Kawerau 1985). Additional confirmation is provided by the Acts of Syrian Martyrs (Assemanus, I, 15), which says that widespread persecution of the
1879, 14). Moreover, this date
is

Christians began in the 31st year of the reign of Shapur II (=340/341), which

is

equated

with the

1

17th year of the empire (=223/224).

228

Michael Airam

ARDASHIR'S EASTERN CAMPAIGN

229

13

14

17

19

20

21

22
10
11

23

24

12

Kuns this torisches Museum, Vienna

Ardashir's coinage can be easily divided into three major phases according to images and legends (Alram & Gyselen 2003).

its

PHASE 1

is

represented by a single type of coin minted in series (drachm, half-drachm,

1/6-drachm), which presented Ardashir as the

new king of

Fars (Fig.
there
is

1).

The

types

and

denominations are clearly rooted

in Persian royal coinage,

and

a seamless

transi-

tion from the previous coinage group of Ardashir's brother Shapur.

point of view, Ardashir's reign as king of Fars

From a numismatic probably did not last very long. The mint
main

may be assumed

to be Stakhr (Group/Mint A), which served for centuries as the
Fars.

mint of the kings of

PHASE 2, Ardashir's
periods:

first

minting phase as 'king of kings' can be divided into three
using the old frontal

Period

I

begins with a small bronze issue, which, while

still

portrait of Ardashir

from Phase

1

on

the obverse, already shows the

new fire

altar

on

the reverse (Alram

& Gyselen 2003. type 1/2,
now
includes the

nos 5-7).
fire altar

Period

II already

shows Ardashir as the new king of the Iranians. The

on

the reverse also
is

new explanatory

circumscription 'Fire of
this type,

Ardashir'. It

remarkable that Sasanian gold coinage began with

as yet

known

only from a

single dinar coin (Fig. 2). It is tempting to

which is assume that this
fire altar

was a

festive issue to

mark

the lighting of the royal

fire for

Ardashir.

On this first gold
on the

issue Ardashir bears only the title of 'king of the Iranians',

and the

reverse

is

not yet decorated with the diadem of the great king.
III

Period

of phase 2 changes

this:

the titulature

is

now enlarged

to 'king of kings

lire altar is enwreathed by a diadem band (Figs 3 and 4). Based on typological details on the obverse, such as the pectoral star on Ardashir's robe (B), the ribbed diadem bands (C) as well as the varied spelling of the obverse legends Cry n / "yr n) (according to Skgerv0 in Alram & Gyselen 2003, 46), the bulk of

of the Iranians' and the

282

G.

A. Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE

283

The situation is rather different with regard to
period.

the materials relating to the Sasanian

Apart from Gyaur-kala and Erk-kala mentioned above we are also familiar with the fortifications at Chilburj (Gaibov et ah 1990). In some important respects the fortifications at Gobekly-depe and Chilburj are similar. In both cases they were erected on a 'foundation' provided by the ruins of the Parthian fortifications, and towers played a significant role in their system of fortifications. The walls at both sites have central corridors, which are divided up into interconnecting sections. All this enables us to assume that there existed at that time in the Merv oasis a number of generally accepted principles for the building of fortifications, which were used in centres of population of
various kinds.

BADER,
In

GATBOV, V. & KOSHELENKO, G. 1995. 'Walls of Margiana'. In A. Invernizzi Papers on Central Asian Archaeology in Antiquity, Florence. the Land of Gryphons
A.,

(ed.),

DURDYEV, D.

Kishman (Report on excavations carried out in 1956-57)', Trudy Instituta istory; arkheology i etnografy Akaclemy nauk Turkmenskoi SSR (Proceedings of the of the Institute of History, Archaeology and Ethnography of the Academy of Sciences SSR) V, Ashgabat. Turkmenian GAIBOV, Y., GUBAEV, A. & KOSHELENKO, G. 2002. 'On certain features of the official ideology
1959. 'The City-site of Old

of Parthian Margiana', Kulturnye tsennosti (Cultural Values) 2000-1, St. Petersburg. NOVIKOV, S. 1990. 'Chilburj', Bulletin of the Asia Institute, GAIBOV, V, KOSHELENKO, G.

&

New
fortifications applied only to

Series 4.

However, the similarity between the two systems of

Gobekly-depe in the early Sasanian period. The subsequent development of these sites was characterized by a dramatic simplification of their defensive systems and, therefore, to a drop in their effectiveness. It is worth pointing out in this connection that in the last stage of the walls at Gobekly-depe they ceased to perform a defensive function.

GOBL, R. GUBAEV,
Delta.

und Chronologie der Miinzpragung des Kusameiches, Vienna. (eds) 1998. The Archaeological Map of the Murgab Rome. Preliminary Reports 1990-1995,
1984. System

A.,

KOSHELENKO, G. & TOSI, M.

KOSHELENKO, G. A. 1963. 'Parthian Fortifications', Sov. Arkh. 2. KOSHELENKO, G. 1996. 'Bullae from Gobekly-depe. General Problems
M.-F. Boussac
leniya

and Main

Subjects'. In

The

corridors within the walls were divided into accommodation 'cells' consisting of 2-3 rooms and, moreover, doorways were cut through the inner wall to lead into these
'cells'

& A. Invernizzi (eds), Archives et Sceaux du monde hellenistique, Paris. KOSHELENKO, G. A., GUBAEV A. G., GAIBOV, V. A. & BADER, A. N. 1995. Sistemy rassei

irrigatsiya v

Mervskom

(Fig. 14).

of settlement and irrigation in the

oazise (Turkmenistan) ot epokhi bronzy do srednevekovya (Patterns Merv oasis [Turkmenistan] from the Bronze Age to the

Medieval Period), Moscow.

KOSHELENKO,

NIKITIN, A. B. 1991. 'Coin finds and questions of stratigraphy at Mezhdunarodnoi assotsiatsy po izucheniyu kultur Tsentralnoi Azy (Bulletin Gobekly-depe', Bulleten of the International Association for the Study of the Cultures of Central Asia) 18. KOSHELENKO, G. A. & USMANOVA, Z. I. 1964. Towards a history of the fortifications in the Cities of ancient Merv', Izvestiya AN TSSR (Reports from the Academy of Sciences of the
G. A.

&

Turkmenian SSR), Social Sciences

Series

1.

LOGINOV, S. D. & NIKITIN, A. NOVIKOV, S. V, NOVIKOVA, O

B. 1986. 'Coins with the

G.

& KOSHELENKO, G. A.

the settlement of Gobekly-depe', History, No. 6. University), Series 8

horseman from Merv', Sov. Arkh. 3. 1994. 'Levels from the Yaz-III era in Vestnik Moskovskogo universiteta (Bulletin of Moscow

PILIPKO,

V N.

PUGACHENKOVA,
Istory 2.

1980. 'Parthian bronze coins with the sign ? under a bow', Vestnik Drevnei Istory 4. G. A. 1952. 'Parthian fortresses in Southern Turkmenistan', Vestnik Drevnei

PUGACHENKOVA,

G. A. 1958. Puti razvitiya arkhitektury Yuzhnogo Turkmenistan pory ifeodalizma (Paths of development in the architecture of Southern Turkmenistan rabovladeniya during the slave-owning and feudal eras), Moscow. PUMPELLY, R. 1908. 'Description of the kurgans of the Merv oasis'. In E. Huntington,
Explorations in Turkestan
1,

Washington.

TASHKHODZHAEV, S. S.
Ashgabat.

1962. 'Section through the city wall of Gyaur-kala', Trudy

YuTAKEXll,

TOLSTOV,

S.

M.
Z.

1948. Drevnei
I.

Khorezm (Ancient Chorasmia), Moscow, 91-3.
Trudy

USMANOVA,

1989.

A Section through the Fortified Wall of Erk-kala in Old Merv',

YuTAKE XIX, Ashgabat. A. 1999. 'The fortifications of Erk-kala and Gyaur-kala', Kulturnye tsennosti ZAVYALOV,

V

(Cultural Values) 1997-8 Issue.
r

igure 14.

Room with

large vessels, built within

an

earlier intra-nmral corridor.

280

G

A. Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE
when the eastern section
the towers

281

protruding buttresses on the outside, two of which were found of the southern wall was cleared (Fig. 11),

no longer protruded, and that the corner towers

for all intents

and purposes

After a time the walls were rebuilt: the upper part was destroyed and replaced by a new one but in accordance with the old principles (two walls and a central
corridor)

disappeared during the last stage of the existence of the fortress (Fig. 13). To sum up we can say that the excavations at Gobekly-depe have provided material the study of the history of fortifications of Margiana during the Parthian and
for

and using

the traditional building methods.

which were from each other but linked by narrow doorways. In the outer wall there were arrow-slits: remains of these were recorded in two places. The next reworking of the walls at Gobekly-depe was less radical. The outer wall was twice strengthened by an additional skin of mud bricks, as a result of which its overall width was increased to approximately 5.6 m. Naturally, the arrow-slits lost their importance, and the warriors of this garrison were only able to take up their positions on top of the wall (Fig. 12). Furthermore, this widening of the outer wall meant that
not, however, completely isolated

imately 9 (the outer wall measured 2.5 wall 3.5 m). This wall was cleared over a long stretch enabling us to record its structure. First and foremost, the corridor within the walls was divided into sections,

m

The overall width of

m

the walls was approxin width, the corridor 3 and the inner

m

Sasanian periods. The materials relating to the Parthian period are particularly signifiMerv Gyaur-kala and cant. Gobekly-depe is the only site in Turkmenistan (apart from where we have a fairly reliable idea of the character of fortifications thanks Erk-kala),
a starting-point for subsequent to completed excavations. These materials can serve as investigations. It would appear to us that, at this point in time, it would be somewhat

premature to draw any kind of general conclusions regarding the nature of fortification in Parthian Margiana.

^WWfSWMT

II

^0^^%

*mmm
t

-mmmm.

01 2

54- 5 6

78910m
T,

Figure 12,

Three main stages in the rebuilding of the walls of Gobekly:
Sasanian.

Parthian, and II and

III,

Figure 13.

Fortifications at

Gobekly in the Sasanian period, the

final stage.

278

G.

A. Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE
The
wall.
fortifications

279
their lay-out

were also rebuilt at

this time.

The idea behind

was

retained with defensive walls

on three

sides

and an entrance

in the middle of the south

upper parts of the walls were destroyed, the corriwithin the walls and the internal space of the towers were filled with bricks and dor clay. The new walls and new towers were built on this 'foundation' (Fig. 10). The state of preservation of the towers was, unfortunately, poor. We assume that the original

As

in the central building, the

form was retained, but, unlike the towers of the Parthian period, there were no internal rooms but they were monolithic. The warriors defending the towers would therefore
have had to stand on the surface at the top.

and repaired several times. The was erected immediately over the in-filled remains of the Parthian wall and also consisted of two walls with a central corridor. The outer wall had a base of several bands ofpakhsa, laid on the remains of the Parthian wall and on debris from that wall lying in front. The thickness of this pakhsa 'cushion' was approximately 2 m, of outer wall, 1.75 m while the wall itself measured 7 m in width, made up by 2.75

The Sasanian

walls of Gobekly-depe were rebuilt

early Sasanian wall

figure 9a-c.

Three stages

in the life

of the Sasanian building at the centre of Gobekly.

of inner corridor and 2.5

m m of inner wall. Unlike the Parthian wall, the Sasanian wall

was made of

mud

bricks (of approximately the

same dimensions). There were some

igure 10.

Fortified walls

from the Sasanian period,

built

above the Parthian

walls.

Figure 11,

Buttress of the Sasanian wall.

276

G.

A. Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE

277

Figure 7.

Arrow-slits in the Parthian south-east tower.

the rest of the wall.. This, of course,
in breaking

made it easier

Figure 8.

Building erected near the fortified gates of Gobekly at the end of the Parthian period.

to resist enemies, if they succeeded

through the gate. Another way of strengthening the defensive potential of the fortress was by making the doorways within the wall extremely narrow, approximately 2 m in the outer wall and even less in the inner one.
the Parthian fortifications at Gobekly-depe were in keeping with the highest standards of fortification techniques of the time. The walls were fairly high and
It is likely that

rebuilding

began with the central mound. The roofs and upper parts of the walls were

destroyed and the

with rubble with considerable care. The bricks left over from the destroyed parts of the earlier building were laid out in more or less even rows and the gaps between them filled with clay, so as to ensure that the new building should

rooms

filled

The thickness was to a large extent due to the use of an internal corridor, which turn would have reduced the costs of labour and materials required for the construction of the wall. The basic features of the defensive system were the substantial towers and the upper walkway on the wall. The unusual structure of the arrow-slits
thick.
its

have a firm foundation.

in

This building was not as well built as the Parthian fortress, and its plan was unusual. The centre was in the shape of a not quite regular monolithic hexagon, surrounded on The five sides by a rather narrow corridor with an open area on the sixth, south side.
central part

made it possible for a single archer to fire arrows in three diagonal directions.
three rooms,

It

should

be noted that at the very end of the Parthian period a small building, consisting of was erected against the gate tower (Fig. 8). This reduced the defensive potential, but it is likely that the times were quiet.
In the Early Sasanian period the fortress changed dramatically. There is no evidence that it underwent any violent demise: Gobekly-depe was simply abandoned. Some time later a different group of people settled among the ruins and rebuilt them.
to

of the building was surrounded by subsidiary rooms, in some of which there were large vessels (khums). Unfortunately we do not know how the upper part of the walls was arranged or the central hexagonal platform itself, since this part of the

show

from the erection of the triangulation station, set up at this point. All that should be noted is that the narrow corridor was filled with ash. It is unlikely that this ash would have been the remains of a fire: it is probable there would have been
building has suffered

an
to

To judge from

the coin finds the interval between these two periods

was

short.

The

on the platform. The building was repaired a number of times: the entrances some rooms were closed, and a wall built in front of the platform (Fig. 9 a-c).
altar

274

G.

A, Koshelenko

Figure 5.

Exterior of the Parthian south-east tower.

Figure 4.

Parthian Gobekly, construction of the fortifications (alternating layers of bricks and
pakhsa).

towers had been built in two stages with a break between the two. The lower part of the tower room had been filled with densely packed rubble. In the walls of the tower there were several arrow-slits. They were of a fairly unusual design: each arrow-slit consisted of three parts, i.e. the archer could use any of the three slits: they sloped downwards
it was possible to fire arrows at the probable that not only the rooms inside the towers but also the area (with a parapet) on the roofs of the towers were used for defensive

(Fig. 7).

The towers protruded
bottom of the

several metres, so

enemy

at the

walls. It is

purposes.

The system for defending the gate is also of interest. There was one tower to the east gate, but only a small protruding buttress on the opposite side. Access to the tower was only possible via the corridor within the wall. Its state of preservation was such, that we do not know whether there were any arrow-slits, although we assume that it had been constructed in the same way as the corner towers. It is essential to point out
of the
that the fortifications in the gate

area— part of the corridor within the wall, two door-

ways

in the walls

and the tower

at the gate

— were cut

off by

two transverse walls from

Figure 6.

Interior of the Parthian south-east tower.

272

G.

A. Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE
2)

273

It

During the Parthian period, a small fortress was built at the edge of the oasis (Fig was almost square in plan (100 X 100 round the top of the walls). A

m

building (45
building

X

rectangula

37.5

m) 5 was

built over the ruins of the

the walled area.

A distinctive aspect of the building is that
a rule were vaulted.

was made of mud bricks measuring from 40 X 40 X 10 cm
ceilings as

Yaz III manor in the centre of it had only one entrance. The
to

42 X 42 x

12-1

cm. The

variety of functions. It would appear to have been a residence for the state dignitary in charge of the fortress— in the centre of the building there are rooms to accommodate his family. 6 At the same time, however, the building served as a 'warehouse', to which specific commodities were sent from Merv and from which

The fortress served a

were then distributed further. 7 The warehouses and those who worked in them, and also the dignitary's servants, were accommodated in small rooms, which surrounded
the

thev

centre of the building.

The excavations
fully excavated

at

Gobekly-depe provided a

fairly detailed picture

Parthian fortifications.
as well as the gate

The south-eastern corner tower and
sections of the wall

and— to a lesser extent —the north-east and
and two

of the nature of the tower at the gate were
south-west corner towers

on

the south side.

We can assume

that

the fortifications were erected in two stages. This can be clearly seen from the south-east corner tower and the section of wall, which adjoins the gate on the west side. The constructions relating to the later stage were a few centimetres wider than the earlier ones, as can be discerned in the sections. We have no definite proof, but nevertheless
the

impression emerges that, after the beginning of construction work, there was a pause and that construction was only completed after this pause.

The plan of the fortifications is as follows: there were towers in all four corners and an entrance in the middle of the south wall, which was also protected by a tower located to the east of it (Fig. 3). The Parthian wall had an overall width of approximately 7 m: there was an internal corridor about 3.5 m wide. The outside wall was slightly wider than the internal one: while the former was 2.2 wide, the latter was only
t

Figure 2.

Plan of Parthian Gobekly.

ruvtrui

rtrtrtrtn
,
,

vumrt/umruv

!WU~y]

1.3

m m wide. The surviving height of the wall varies from 3.6 m to almost 6.0 m. It can

be assumed that the ceiling of the corridor was vaulted. The walls were built as follows: a layer of pakhsa was laid on a surface of closely rammed earth, with alternate layers of mud bricks (of the same size as those used to erect the central building) and pakhsa

above (Fig. 4). It can be assumed that a wall constructed in this way already constituted a serious defensive system, yet there were also arrow-slits in the wall itself. The warriors who defended the walls must have taken up their positions on the upper part of the wall,

10

5

o

10

20

30

40m

Figure 3.

Parthian Gobekly, reconstruction of southern facade.

5

It

should be pointed out that this building

is

the only building from Parthian Margiana, which has been fully

excavated.
6
7

behind the parapet. Particular importance was attached to the towers in this defensive system. The fully excavated south-east tower (Figs 5 and 6) and the two other partly
excavated towers were constructed in a similar way.
the wall: alternate horizontal layers

An interesting detail of the interior is provided by the fireplaces, similar to those in the houses at Ai Khanum.
The numerous
bullae found here

would point to

this function, see

Koshelenko 1996, and Gaibov below.

The same technique was used as for of mud bricks and pakhsa. Like the walls, the

270

G. A.

Koshelenko

THE FORTIFICATIONS AT GOBEKLY-DEPE

271

1989). Unfortunately the impressive results of the excavations in the south-eastern corner of the fortified walls of Gyaur-kala, undertaken b Y the International Merv Project, have not yet been published in full (Za'vyalov 1999)

and Erk-kala (Usmanova

concerned with the results of research into the fortifications of one of Merv Oasis— Gobekly-depe This research was part of a wider programme concerned with the compilation of an archaeological map of the Murghab Delta, carried out by a joint Russo-TurkmenItalian mission (Gubaev et al. 1998). In addition to reconnaissance work and the recording of monuments discovered in the course of this project, the programme
is

This paper

the

monuments

located at the north-western edge of the

some sites. These were aimed to verify the results of reconnaissance connected with the dating of monuments, their dimensions and so on One of the sites earmarked for excavation was Gobekly-depe, 2 which, on the basis
of work undertaken by YuTAKE, had already been defined as a small' town of the Early Parthian period (Pugachenkova 1958, 44-5). Our research subsequently made it possible to gain a clearer picture of the nature and history of the monument. 3 As a result of excavations carried out over a long period, it became clear that there were
three stages in the history of the
fortified
site. During the first, Yaz-depe proved to be a small manor, situated near the centre of the Early Iron Age oasis (Novikov et al 1994). At the beginning of the Hellenistic period this manor, like all centres of popu-

also allowed for excavations at

lation in the northern part of the oasis,
first

was abandoned. The site was reoccupied in the century bc or perhaps even in the first century ad. It was precisely at that time that the previously abandoned northern part of the oasis was rapidly and intensively
settled. It can be assumed that this settlement proceeded in an organized rather than a spontaneous way, on the initiative of a central authority, under its control and,

Figure

1.

Excavations at Gobekly: aerial view.

apparently,

expense of that authority. Large new irrigation systems were created and new centres of population built (Koshelenko et al 1995). This was a grandiose undertaking, the scale of which was to exert an enormous influence on the subsequent history of the oasis. Many new centres of population were settled, including Old Kishman, New Kishman, Chilburj, Durnaly, Kyrk-depe and many others. Most were fortified and had a regular plan. It was during this period that Gobekly-depe came into existence
(Fig. 1).

at the

to the early third century

AD

(Pilipko 1980).

Coin

finds relating to the

Sasanian period

were slightly more diverse and included bronze coins with a depiction of the 'Merv Nikitin 1986): these belong to an issue of the vassal king of horseman' (Loginov

&

4 There is one Kushanshah coin of Bahram II, and coins which appear to have been issued at the local mint by some of the first Sasanian kings up to the reign of Shapur II (309-379 ad). There are no later coins. It should be

Merv during

the rule of Ardashir

I.

coins were found during the excavation of Gobekly-depe: these are distributed in the various levels in such a way as to make possible a fairly reliable dating

Numerous
site

pointed out, by the way, that,

when a

wall

was built in the
site

sixth century in the north of
it

the oasis (usually referred to erroneously as the 'wall of Antiochus'),

passed to the

of the

two distinct periods, the Parthian coin finds of the Parthian period there are only 'copper drachms' originating from the local Merv mint and dating to the first through and the Sasanian.

(Koshelenko

& Nikitin
the

1991). There are

south of the ruins of the then abandoned

of Gobekly-depe (Bader

et al. 1995).

Among

many

Gobekly-depe are known: it lasted from the first to the late fourth century ad. Furthermore, coin finds make it possible to divide that period into two fairly distinct stages, Parthian and Sasanian.

Thus the

precise dates for

2

1908,223^1.
3

This monument was discovered by Pumpelly and designated as a burial-mound of a 'transitional type' Pumpeflv v v
'

''

Hi

accordance with the system devised by R. Gobi (1984, 85,

pi. 116,

nos 1096, 1097),

it

dates approximately to

the eighties of the fourth century AD. It should be noted, however, that the dates given

by Gobi for Kushano-

Several short reports

on the excavations have already appeared: the most recent

is

Gaibov

el al. 2002, with

bibliography.

Sasanian coins do not appear convincing to me. Yet, whichever dating system is used, this issue cannot be later than the time indicated here. Thus, the latest date for Gobekly-depe remains the same.

Academy of Sciences, Moscow

Different types of fortifications
logical investigations, fortifications
villages

are

an important

historical source, informing us

about the character of the societies in question. Naturally,

when

it

comes to archaeocities,
1

used for the protection of state borders,

and other types of centres of population are an important focus of
in the

study.

When

the South Turkmenistan Archaeological Multi-Disciplinary Expedition

(YuTAKE) began work

Merv

Oasis in the late 1940s, special attention was, of

course, paid to the nature of fortifications of the ancient centres of population,

and

many were recorded (Pugachenkova Karakum Canal had not been
did not exist and bulldozers
built,

1952; 1958, 43-59). Their state of preservation,

particularly those in the north of the oasis,

was incomparably better than today. The the water-table was low, many of today's villages

had not flattened small tells to create cotton-fields. After the Canal was built dozens of monuments, particularly small ones, were destroyed. Most of the larger monuments survived, but the surrounding archaeological landscape had disappeared without trace and the monuments themselves were at risk. Because of the rising water-table, they soon lost structural features, which had been clearly visible, and degenerated into shapeless mounds. One objective of researchers working in the early years of YuTAKE's existence was
study the fortifications of the various centres of population in the oasis. Unfortunately, their conclusions about both the nature of the centres themselves and
to

of their fortifications were drawn only on the basis of visual surveys and collections of surface finds. While many conclusions still appear to be justified today, some have proved to be inaccurate (Koshelenko 1963). Later on, YuTAKE teams stopped working on
small-scale
fortifications in the north, and only work (and, as far as we can tell, mainly incomplete projects) was undertaken by staff members from the Institute of History, Archaeology and Ethnography

in

in the city-sites

Ashgabat (Durdyev 1959). YuTAKE excavations were only carried out in Merv itself, of Gyaur-kala (Tashkhodzhaev 1962; Koshelenko & Usmanova 1964)

1

This point was stressed by Tolstov (1948, 91-3),

whom we can single out as the 'pioneer' in the study of ancient

fortifications in Central Asia.

312
could prevent the

Olivier

Lecomte

fall of a state which, like many previous centralized powers, had already started to disintegrate because of internal pressures and dynastic problems. To these causes should be added the fragility caused by the immensity of the territories

controlled.

References

ARRIAN.
by
Fr.

Arrianus fragmenta scriptorum de rebus Alexandri

M.

Arriani Anabasis et Indica, edited

Dubner.

No date.

Firmin Didot,

Paris, 83.

ATAGARRYEV, A. E. 1994. 'Kerpt itch lining syry ajanboijar'. Turkmen Medenieti 2, 6-8. BOUCHARLAT, R. & LECOMTE, O. 1987. Fouilles de Tureng Tepe. Sous la direction
Deshayes. 1 Les Periodes Sassanides et Islamiques,

V.
de Jean

A.

ZAVYALOV
of Material
Culture, St. Petersburg

Institute for the History

ERC,

Paris.

BRIANT,

P.

1996. Histoire de {'Empire Perse, Fayard, Paris.

et al. (eds), The Cambridge Ancient History, vol. 4, Cambridge (2nd edition), 165-93. GYSELEN, R. 1987. 'Bulles et sceau sassanides'. In Boucharlat & Lecomte 1987, 187-91. HANSMAN, J. 1968. 'The problem of Qumis', JRAS, 11 1-39. HARMATTA, J. 1996. 'The walls of Alexander the Great and the Limes Sasanicus', Bulletin of the

CAH IX. 1966. The Roman Republic 133^t4 bc, 575. CHRISTENSEN, A. 1971. L'Iran sous les Sasanides, 1944, reed., Osnabriick. FRANCFORT, H.-P. 1988. 'Central Asia and Eastern Iran'. In J. Boardman

Introduction

The historic urban centre of Merv is
Turkmenistan. During the Achaemenid
within a polygonal enclosure wall, which

located in the

Murghab

oasis in

modern day

period the
still

first city

(Erk Kaia) was founded

stands to a height of more than 20

m and

occupies an area of
Kala),

some 12 ha

(Fig.

1).

A massive city,

Antiochia Margiana (Gyaur
citadel in the middle of

Asia Institute 10, 79-84.

was

built in the Seleucid period:

Erk Kala formed the

HERZFELD,
I,

E. 1924. Paikidi

monument and inscriptions of the

early history

of the Sasanian empire
Isidore

Berlin.

HUFF, D.
JUSTI,

1981. 'Zur Datierung des Alexanderwall', Iranica Antigua
9.

XVI, 125-39.

of Charax,

Parthian Stations. W. H. Schoff (ed.). London, 1914, 1976.
F. 1963. Iranisches

KIANI, M.

Y. 1982. Parthian Sites in Hyrcania.

Namenbuch, Marburg, reprinted 1963. The Gurgan plain,

Berlin.

KHLOPIN, I. N. 1983. Istoricheskaia geografiia iuzhnykh oblastrei Srednei Azy, Ashkabad. PAULY WISSOWA 1914. Paulys Realencyclopadie der Klassischen AltertumsWissenschaft, Neue Bearbeitung begonnen von Georg WISSOWA. A. Driickenmuller. Stuttgart.
POLYBIUS.
1960.
classical library

Anabase d 'Alexandre, with an English translation by W. R. Paton, London, Harvard University Press, Cambridge, 1960, 175-9.
1961. Histoires, Texte etabli et traduit par

vol. IV,

Loeb

QU1NTUS CURTIUS.
Lettres, Paris, 1, 178.

H. Bardon, 2nd

ed.,

Les Belles

of Merv. The Achaemenid city of Erk Kaia became a citadel of the Hellenistic city of Antiochia Margiana (now known as Gyaur Kala). The later Islamic city of Sultan Kala developed to the west. The location of the excavation of the defences of Gyaur Kala, the subject of this
Figure
1.

The

cities

paper,

is

shown

inthe south-west of the

city.

Proceedings of the British

Academy

133, 313-329.

© The British Academy 2007.

310

Olivier

Lecomte

GORGAN AND DEHTSTAN

311

progressive infiltration of played the role of obligatory passage in order to control the beginning to settle over all the territory of Hyrcania. populations who were Originating themselves from Dehistan and the steppe world, the Parthians soon

BC

power. understood the best strategy to observe populations who were hostile to their second line of fortification inside This is why they built Alexander's wall, its forts and a mentioned above, like the Roman limes. This policy seems to Gorgan, something, as I the Parthian and the Sasanian, have functioned well for both the centralized empires, caused the final collapse of the but the alliance of the Turks and the Hephtalites
Sasanian empire. Neither Derbent's nor Alexander's wall and
its

related forts

and

sites

BD

BE

BF

Figure 10.

Geoktchik Depe: plan of the gate of the

late

Sasanian

fortified

farm.

"f
quite different for the material culture of the Tchols which is homogenous from the Uzboy in the north to the plain of Gorgan the south. This observation indicates that the construction in the Parthian period and the restoration of Alexander's wall at the end of the Sasanian period,? far from blocking the advancing tide of nomads at the northern frontiers of the empire, rather
is

Gorgan The

reality

clearly identified as

m

5m
'Contra Harmatta 1996, agree with Huff's dating of Alexander's wall and believe that natural sciences-i 1 e the study of the Caspian Sea level-are reliable for this particular purpose, Huff 1981.
Figure 11.

Geoktchik Depe: plan of the 'manor' of the

late

Sasanian

fortified farm.

308

Olivier

Lecomte

GORGAN AND DEHISTAN
t

309

Furthermore as already mentioned, several Sasanian rulers used Hyrcania as a Place from which to launch military expeditions against the nomads. This situation could have contributed to the creation of a sort of no man's land in the south of Dehistan, pushing populations either to move to the north-west of the plain, or to keep away from the border to the south. Among the peoples who motivated these defensive a Wer <Whke HunS kn ° Wn fr° the fourth <*"**. ' °: lithvTttl : Ir tle by httle established themselves in the territory of Hyrcania finally to dominate it including the plain of Gorgan. It is against the probable descendants of the Chionites' known in the fifth century as 'Tchols', that the Arab warriors were to conquer he 4 region in the eighth century.

—H-

r

w°f"

m

^o

SasaDian-isiamic selUemcnt

Area

I

(gate)

A typical settlement of that period in Dehistan is Geoktchik Depe, a fortified farm occupying 4.5 ha The enclosure was constructed to the east of the largest, Iron Age tepe (Fig. 9), probably the eighth century ad. It measures 223 in length by 206 m width with walls 2.60 thick, constructed of square mud bricks (0.46 X 48 m) drCUlar tOW6rS in the C ° rners and semicircular'towers egularly spaced along the walls. The access is on the east, opposite the main tepe. Unlike major Sasanian sites and some contemporary sites, which were highly structured and densely occupied, the enclosure of Geoktchik illustrates a plan which is quite different but widespread in Dehistan. Scattered over a surface of four and a half ha the structures are few in number (five), well-spaced from one another and aligned approximately north-west/south-east leaving a large space free on both sides Two excavations were opened in 1997 by the French-Turkmen Archaeological Expedition: ihe gate {Fig. 10) is composed of two massive three-quarter towers, generating a large protruding single round tower protecting access to a porchway, which opens onto habl vaulted as indicated by the us brick fragments on the floors (which were themselves ? paved by fragmentary mud bricks) The square mud bricks measure 0.46-0.48 per side and are 0.15 m thick. Some buildings

m

m

m

m

™T?

VT^u ^

^

z~^r«
v,

,t

r

'

™°
ainst

n° WeSt U a the t^flTluZ ^H aCtCnStlC Plan ^the P rotrudin S tower mayP be*of Arab ™* / ° am a f A. Northedge personal communication) and does not permit a
Cieared n
rth "
buiIt
'

^
1

m

f

^

origin

Figure 9.

Geoktchik Depe: the Iron Age circular mound and the

late

Sasanian

fortified

farm.

date for the construc-

be contemporaneous with Tureng Tepe's temple and domestic dwelling, whose pottery is identical with that of

^71^71 7 n fy would i before 716-717. Our farm Z then
The central mound (Fig.
1).

Xnl^ Sn 7T

fmmti

re the inVaSi ° n

e

° f thC Plain ° f G ° rgan under the cali P hate of th dt the C ° nqUeSt b the Arabs was * «>t completed
'

in this period,

fire

Dehistan
(1 .50

an agricultural region with a sedentary population, but also a nomadic one, as indicated in the historical sources. This fortified farm where pottery was also produced, indicated by numerous kiln fragments visible on the surface, most certainly

d -ore than 200 two constructed hearths were cleared. The few remains recovered allow us to date this structure to the seventieth centuries ad This type of fortified establishment, for agricultural as well as defensive purposes (it would serve time of danger as a refuge-enclosure, explaining the limited surface area occupied by the buildings), represents the synthesis of the occupation of Dehistan

dWelKng Ln 200^ °TH rooms and "^^ " ^^ T m Three
enCl ° Sed

Partial excavation of the largest

mound

Place

<™ —

m high)

characterized the settlements of the Tchols, descendants of the Chionites, themselves related to the Hephtalites. The open space in the interior of the enclosure wall could

accommodate yurts and herds

in time of danger, as well being a place of exchange.

m

the fifth century onwards, the determining element in the plain of Misrian have been the presence of the Tchols, for whom Dehistan was most certainly seems to the southern limit of their territorial extension. By territorial extension I mean effective control of a region whose capital lay almost 200 km to the north-west of the plain of

From

306
.22*

Olivier

Lecomte
* Arcliak Dchistatt siws APdriliiasiiUitJuasauiaiisiws

GORGAN AND DEHISTAN
century, the

307

Khagan

lived at

Balkan6

to the east of the bay of Krasnovodsk'.

That

is

edge of Dehistan. to say the very northern conquered land, he (Yazdgard II, 'In the

438^57) founded

the city of Sharestan~i

to be near the frontier exposed to Yazdgard where he lived for several years in order 1971, 287, n. 1). Here we have proof of the devastation by the barbarians' (Christensen conquest or maintenance of order are regof instability of Dehistan where operations

of the region, at least temporarily. ularly necessary to (re)take control
is

This hypothesis

these farmers were the auxiliary supported by the fact that: 'Much more useful than peoples who lived within the limits of the empire troops provided by various warlike

(Christensen princes, had a privileged position.'. but who, governed by indigenous the empire' in the north clearly that the 'limits of 1971, 209). This last text indicates
. .

were nothing less than ensured. uncertain status of Dehistan To answer at last the question above, it is clear that the system could not depend on the authormeant that the maintenance of the irrigation maybe, for the short period when one of them made a town ity of the Sasanians, except, The total absence of Sasanian sites in the no man's land in Dehistan his main residence. wall and north-west Dehistan implies, which made up the zone between Alexander's latter, the maintenance of the mam canal nevertheless, if water was to be brought to the towns and farms of the northern plain between its source at the Atrek and the fortified in this vast enterprise by the different of Misrian. So, the necessity of collaboration all is evident. It is clear human groups whose farming practices ensured survival for probably at the level of tribe co-ordinated by the local elites,
that this

work had

to

be

and

clans.

Of

course,

some kind of recognition of

the Sasanian political

power took

the irrigation system which it controlled place in this process, since the latter could close in today's Iran. from its water take on the Atrek, near Tchat, thickness of Uhi Kizylli, distinguished by its size, its plan, and by the

The

site

of

QLdel:

10 km

its walls,

esis,

Figure 8.

Dehistan: distribution of Early Iron Age, Sasanian and late Sasanian

sites.

earlier

propose, as a hypothcould have been a garrison city, a royal foundation. We As I mentioned, in this period as in to identify it with Shahrestan-i Yazdgard. south and north of Hyrcania, the ones, the material culture is the same in the

architecture

and pottery assemblages being

directly comparable.

At

Kirpitchli in northern Dehistan,

Turkmen

archaeologists discovered a clay seal

depended on an agricultural economy and pose the problem of canal in fact, been discovered in the south over about 60 km. How and in what socio-economic and political framework was the main canal maintained? We know that the plain of Gorgan was part of the Sasanian empire from the third century, but not Dehistan. The historical sources are clear on this subject: 'However, Yazdgard had inflicted a defeat on the king of the Hunnish horde of Tchols, who lived .' in the north of Gurgan I emphasize in the north and not to the north of Gorgan, understood as Hyrcania (Christensen 1971, 287). 'From the second half of the 5th
sites

These

maintenance.

No site has,

(Atagarryev 1994; Gyselen 1987). Its impression identical to one found at Tureng Tepe official of the empire, 'defender and judge legend mentions the name of a government provides at least the possibilof the poor', practising at Gorgan. This seal impression Hyrcania in the late Sasanian period, whenever it ity of some administrative unity of This context made possible was thought necessary to integrate Dehistan into Hyrcania. system essential to the economic the maintenance and even extension of the irrigation
survival of the farming

communities

in the north-west.

.

.

*

It is

tempting to relate

river

Uzboy, not

far

of the rocky escarpments of the from the present-day Nebit-Dagh.
this

name

to that

Little

and the Great Balkhan, near

the

304

Olivier

Lecomte

GORGAN AND DEHISTAN

305

.ai^fSmhra^^BpfS^ta^

-

-.

r-

;

V
tw

Figure 6.

Tureng Tepe. The

mud brick patterns on the northern facade (from Boucharlat &
Lecomte
1987).

However, the major Sasanian structure was, as in the last period, Alexander's wall. This was reactivated during the Sasanian period, probably when Hyrcania was used as a platform for military expeditions against northern groups threatening the border. Khosrow I (53 1-579) is known to have played a major role in the protection of his borders. He is also known as a city builder. He reinforced the fortification system by ordering the building of defensive structures against Choi and Hephthalite raids. He. was conscious of the strategic location of the Gorgan plain, as a gathering place for his
armies,

!

!

BffiMBm BhflB
m

and often resided
wall.

there.

So Khosrow

I is the best

candidate for the restoration
38.00

of Alexander's

Archaeology has showed that the material culture of the Gorgan plain and Dehistan

was identical during the Sasanian period, which suggests that the population was the same on both sides of Alexander's wall. In short, those who represented a threat, north of the wall, were settlers, south of the wall and belonged to one single human group. A form of control was therefore necessary to ensure the security of this ethnically and politically 'floating' border. Small forts such as Tureng Tepe, located beyond Alexander's wall, probably played the role of 'police forts', a bit like modern karakuls. Some small forts were also scattered in the plain to control the heterogeneous population.

The distribution of sites in Dehistan confirms this (Fig. 8). The oldest sites represented there, the period of archaic Dehistan, were concentrated in the southern plain. In the Sasanian period, probably in the third-fifth centuries ad, the sites had moved to the north-west and most were established along irrigation canals requiring regular
maintenance. Some, such as Shahdiz Kala, are situated on extensions of the ancient system from the main canal. Compared to the Iron Age, the number of
sites increased,

b'

A.L
\

considerably reduced, perhaps because the fields are less identifiable but also because of the architecture. In most cases this involves small, heavis

but the occupied surface

5

in

Eil&lajss&j
X//7////A

rad'ef en
fariques

pise

towns and farms. Certain groups of buildings have an obvious military aspect. They might be defended by a double wall and ditch, and the most important, such as at Khanly Tepe or Urn Kizylli, would have a citadel dominating a low city.
ily fortified

crues

l^^i^^j

soubassement des
piles

en

pise"

302

Olivier

Lecomte

GORGAN AND DEHTSTAN
10
11

303
16 17

12

13

IS

18

bfc^f

a

5 p

psSriode VI

VZmm periode

Vl(

A8

p6fiodei

1X8

x

9 UMAiHtiaA.A.Pati

<
Figure 5.

General plan of the Sasanian fort

at

Tureng Tepe (from Boucharlat

& Lecomte 1987).

ta

it

decorated with

was protected by horse-shoe towers with a gate opening to the north. Its facade was mud brick patterns at two different levels (Fig. 6). This decoration, as
its

well as the shape of the arrow slots, testify to a central Asian architectural tradition.

There were several building phases during the roughly five centuries of During the last, all the rooms of the inner building were filled with changing the fort into a large observation platform.

existence.

mud

bricks,

At
was

the end of

its

occupation, probably during the eighth century ad, a fire-temple

built inside its fortifications (Fig. 7), together

with a dwelling house.

It is

not

really

Gorgan plain at this time, since the Arab conquest did not occur before the eighth century, and fire temples seem to have continued
surprising to find a Chahar Taq in the

into the tenth century in this area.

300

Olivier

Lecomte

GORGAN AND DEHISTAN
19): 'It is

301

of the first houses of the kingdom .'. And again (Christensen 1971, without reason that the period between Alexander and the accession of
. .

not

trative

and

political integration

of Dehistan into the Sasanian empire of the third

the Sasanians

was termed by the Arab historians that of 'kings of tribes' (muluk'ut-tawd'ij). This term is the translation of the Pehlevi word kadhagh-khvadhdy, 'master of the house', 'ruling prince' ... It is clear that on the northern frontiers of its empire the centralizing will of Parthian power had to content itself with federating some vassals, the chiefs of autonomous clans, 5 who henceforth were the only ones to control these regions.
Dehistan was subject to different phenomena: a reduction of sedentary occupation, a possible return to nomadism, in any case an unstable situation. It is important to emphasize the fact that it is precisely the expanIt is possible that settlement in

century, these texts are evidence that Dehistan was at least culturally an integrated part

of Hyrcania.

Archaeological data
According to recent archaeological work, the occupation of the plain of Misrian was continuous, from the Iron Age to the Mongol invasion.

Data

for the Parthian period are different in the
sites

Gorgan

plain to Dehistan.

dynamic of the Parthians, the centralization of their power in capitals which are farther and farther west, which led to the loss of their control over Dehistan.
sionist

Sasanian and Sasanian-Mamic periods

by a distinchave been recognized in the tive material culture. Pottery, for example consists of four categories: simple ware, grey or black ware, red burnished ware and 'clinky' ware. As for Dehistan, only a single site, Orta Depeslik, has produced pottery datable to the Parthian period by reference to that

Numerous

Gorgan

plain, characterized

of Gorgan.

It is

possible that

we

are confronted in Dehistan with a simple problem of

The

recognition of the material remains.
transition to the Sasanian period
picture, at least for

Such was the case of the Iron Age assemblages, the
in

These present a

documented by archaeology and texts. the beginning of the period, of a situation relatively
is

well

Archaic Dehistan culture, which were only recently recognized to be

use as late as the

close to that described for the Parthian period.

An important piece of evidence for the
,

fragmentation of Hyrcania in the third-fourth centuries (in any case, after the conquest of Gorgan by Bahram II in 284) is provided by Christensen (1971 137) for whom 'The Sasanian empire in the third and fourth centuries possessed a large area in the lands of the north and the east. It included, after the conquests of Bahram II in 284, the following lands in the east: Gurgan (Hyrcania) and all of Khorasan in the expanse contained in this "eastern quarter'". Herzfeld (1924, 37) describes 'The limits of
in the Sasanian period: the south-east corner of the

Achaemenid period. The most impressive
ter

structures of the Parthian period all display a military charac-

which, in a way recalls the

Roman limes.

Such, of course was the case of Alexander's

wall (Fig. 4), a very impressive structure built to control the movements of populations from the north into the Gorgan plain. This wall, known as Alexander's wall (Sad-i-

Jskandar in Persian, KizylAlan in Turkmen), follows the line of the Gorgan river to the north. Before Huff's reinterpretation, this structure was thought to have been built by
the Sasanians.
It

Khorasan

consisted of a huge fortified wall in

mud

brick,

1

70

km in

length
Its

Atrek

.

.

.'.

Caspian Sea, the valley of the The distinction between Gorgan and Dehistan (as part of Khorasan) pro-

defended

vides evidence that at the beginning of the Sasanian period, Hyrcania reduced politically to the plain of Gorgan.

was probably
previous

at regular intervals by function was also to enclose and protect the fields and orchards of the plain. fort of the Parthian period existed at Tureng Tepe, in the Gorgan plain, which

33 small forts

(Huff 1981; Kiani 1982).

and main

A

The

was only

partially excavated

and

is

unfortunately not yet published.

The

structures

social organization

of the ruling

elite

was to be found, as

in the

period, in the privileged situation of the 'seven families', who probably represented the chiefs of the clans closest to the ruling family. Among these seven families (of which three had already occupied this privileged situation under the Parthians), was that of the Aspabadh, which like at least two others descended from the Arsacids and carried the name Pahlav, that is Parthian. The fief of the Aspabadh, which was probably their

excavated there date from the first century bc to the first century ad. Its pottery belongs to the central Asian tradition and, together with that of Shahr-i Qumis, it probably
represents the southernmost diffusion of Yuezhi related cultures.

In the plain of Gorgan, sites of the Sasanian period are widespread. Apart from large settlements, the biggest of which is Asterabad, that is the modern town of

land of origin, was 'Dehistan in Gurgan' (Justi 1963, SpddabatL Herzfeld 1924. Gloss N° 727, cited by Christensen 1971, 105). Although they do not confirm the adminis-

Gorgan, the main component of the previous defensive system, Alexander's wall, seems to have been abandoned at the beginning of the period and replaced by numerous small Tureng forts, scattered in the Gorgan plain, the only excavated example being that of
Tepe.

quite obvious that, for the ruling family, the only way to ensure the support of the leaders of territorially powerful clans was to accord them privileged status in the first circle of power, as, for example, in the case of Gev mentioned above. This policy was also adopted by the first Sasanian rulers.
It is

A small fort was built at Tureng Tepe on top of one of the largest and oldest sites of
Gorgan plain at the beginning of the Sasanian period, probably as soon as Gorgan became part of the empire in 284 ad. Roughly square in plan, 50 m each side (Fig. 5),
the

298

Olivier

Lecomte

GORGAN AND DEHISTAN

299

Sources and History of Hyrcania
Seizing Hyrcania has always been a goal of the neighbouring powers, probably from

of Gorgan). After the end of his reign, the Greeks of Bactria attacked the Parthians and assembled the conquered territories (Tapuria and high Atrek, in particular) into two satrapies. Finally, Hyrcania rejoined the zone of Parthian influence, probably under
the reign of Phraates
the Seleucids
I

Median period onwards, and Herodotus (III, 117) in the fifth century bc even Achaemenids (Briant 1996, 427-8). We know that Alexander the Great, on the way to Herat, passed through Hyrcania (Khlopine
the
describes the taking of Dehistan by the
1983).

(176-171 bc) while the independence of the Parthians from
1

Concerning

this,

Arrian {Anabasis of Alexander III 23, 6) mentioned that he
1 .
.

'advanced towards Hyrcania, to Zadracarta, a city of the Hyrcanians
that this

.*,

and said
is

was

'the largest city

of Hyrcania where the palace of the Hyrcanians
.'
. .

also

found'. Quintus Curtius (VI 5, 1) confirmed the presence of Alexander: 'the king

had

was acquired in 88 bc. Isidorus of Charax (Mansiones Parthicae, §10), carrying out a survey of the Parthian empire on the behalf of merchants, described Hyrcania at the beginning of the Christian era as composed of eleven villages in which commercial outposts were situated. In short, Hyrcania is described everywhere as fertile, densely populated and as an important trading area. We know in total the names of the towns of Talabroke,
Samariane (Saramanna, on the coast), Carta (Zadracarta?), Sirynx (also described as the strongly fortified capital where a Greek colony resided (Polybius X 31, 11)), Sokanda (also on the coast) and the sites of the royal residences of Tape (which some identify at Hecatompylos) and of Tambrax. The latter, like Sirynx, was situated in the
3 west of Hyrcania, and included a royal palace. However, this region quickly became marginal, for the Parthians established a

already penetrated to the ends of Hyrcania

'They had

now

arrived in the city of

Hyrcania where stood the palace of Darius'.

Parthian period
This period is well documented in the

new

and both the plains of Gorgan and Misrian are described on several occasions by travellers and ancient historians. In the fourth to third centuries BC Hyrcania, and in particular the northern part of Dehistan near the river Uzboy (the fossil bed of the Amu Darya), was the cradle of the Parthian Arsacid dynasty, which later installed its capital at Nisa (Mithridatkart) in the vicinity of modern Ashgabat. The Parthians were originally a nomadic people
texts,

Hecatompylos, 4 identified by Hansman (1968) as Shahr-i Qumis in the southern foothills of the Elburz. Apparently it was unstable, not well controlled and threatened by nomadic populations. These nomads represented such a danger that the Parthians were forced to split Hyrcania in two. Gorgan, the southern part of Hyrcania,
capital at

remained attached to the empire, while Dehistan was isolated,
central

left as it

were to the

integrated into the tribal confederation of the Dahistano-Parnis. Parni

is

the

name of
this

the tribe from which the Parthians came; the Daha, the other
federation, could

component of

con-

be the origin of the name Dehistan 'land of the Daha' 2 to designate northern Hyrcania. As Christensen notes: 'The political traditions ... are not abandoned either, when the Arsacids, with the support of leaders who came, as the Arsacids themselves, from the north-Iranian Daha people, and from their levied warriors, establishing themselves first of all in Parthia, created through conquest a new Iranian empire.' 'The Aparnes were a tribe of nomadic Iranian people of the Daha. The founder of the dynasty of the Arsacids was first the chief of the Aparnes' (Christensen 1971, 16 and 220, n. 2). From the third century bc onwards the regions bordering the
. .

Asian world. Indeed the Parthians constructed, in the second-first centuries bc, according to Dietrich Huff, a strong line of defence from the Caspian Sea to the Kopet Dagh in the east. In fact, the effective integration of Dehistan into the Parthian empire during its
phase of progressive centralization towards the west (Nisa, Hecatompylos, Ecbatana, Seleucia of the Tigris) is more than doubtful. As Christensen notices (1971, 19, n. 1): 'Perhaps it is no accident that the country which was the first centre of power in the
dynasty, Parthia (that
this
is

the territory which corresponds to the

Achaemenid satrapy of

.

period more fragmented than any other part of the empire; according to Isidorus of Charax, it was divided into six provinces. One of these six provinces, Hyrcania appeared to have been the hereditary principality of Gev, whose

name) was during

this

had great strategic value for the forces there. Antiochos III in 217 bc dominated the Daha; he then conquered Hyrcania (probably only the plain
Iranian north-east

son or descendent, Gotarzes, mounted the throne. This Gev probably belonged to one

Hyrcaniam progressus Zadracartam urbem petit'. Capital of the province, Zadracarta has been Astrabad or at least in the neighbourhood of the present Gorgan. See Pauly Wissowa 1914, 499.
'In
-

1

identified as

many others. See for example Francfort Wissowa 1914, 514-15. 'According to one tradition, there was (in the Arab period) in Gorgan a city called Dehistan which perpetuated the old name Daha and is found on the edge of the steppe ... but which would have already held this name under the Sasanians'. See also 'Daai', Pauly Wissowa 1914, 1945-6 and CAH
173. See also Pauly

This hypothesis, the most attractive in our view, has been held by

Antiochos III seized these two cities in 210-209 bc. See Pauly Wissowa 1914, 515. 'After having organised the advance as he intended, he reached Tambrax, an unfortified but extensive city which included a royal palace'. The soldiers and inhabitants of the place 'retreated to a city called Sirynx, not far from Tambrax which seemed to be There were three ditches, each defended by a douthe capital of Hyrcania because of its power and location ble row of palisades and then a strong wall. The ditches were rapidly filled and the wall, undermined, collapsed'.
3
.

.

.

4

IX, 575.

28. 7-29.6). city, the Hyrcanians 'killed all the Greeks in the city and then fled' (Polybius century bc Hecatompylos was not part of Hyrcania but occupied 'the centre of Parthia'. Leaving this In the third 28. 7-29.6). city to confront Arsaces, Antiochos III decided to 'penetrate Hyrcania' (Polybius

Before abandoning the

X

X

296

Olivier

Lecomte

GORGAN AND DEHISTAN

297

^^y
^an
fflMis

iMw

Figure

2.

Plain of Gorgan:

Tureng Tepe from the south-east.

£
;y-;:.;^^|.

...

.

.

.

; ""-.^

^•'^^

s
--."
'.

,'•

^"i"'>]^^^^B

HBBiii
''

" r~^s^£|£
.

^-a
•':•-•-

- v
•--

.

.V;".^'.^;-;
.

^•-vrfBilfili '-'*s
3§iS£*»g||

Figure 3.

Plain of Dehistan seen from the Sasanian site of Moqodjik.

f
:

tip
\ \

i

a
\ ]

: :

:::M;.

:

:

|
< s! u

<S

§

8

Tureng Tepe. It is in fact the Bronze Age which best characterizes the plain of Gorgan, with what archaeologists call 'the burnished grey ware (BGW) cultures'. However, whatever the situation was before the late pre-Islamic periods, one should emphasize that from the Parthian period onwards, Dehistan, the cradle of the Pahlavi dynasty, and Gorgan would formally become part of the same administrative entity.
But, as

we shall

see, this ideal

administrative situation does not really correspond to the

textual evidence,

nor

is it

proved by archaeological research.

294

Vasif Gaibov

GAIBOV,
383-8.

V. A.

& KOSHELENKO, G.

depe', Problemy istory, filology, ladtury

A. 1999. 'A Thymiaterion from the excavations at GobeklyVII— Monographs and Articles, Moscow-Magnitogorsk b &
,

GAIBOV, V, GUBAEV, A.

& KOSHELENKO, G. 2002. 'On Certain Features of the official ideology of Parthian Margiana', Kultumye tsennostl Mezhdunarodnyi ezhegodnik—2000-2001 (Cultural Values. International Annual— 2000-2001), St. Petersburg, 191-4.
R. 1976. Die Tonbullen von Takhti-i Suleiman. Ein Beitrag zur spatsasanidischen Sphragistik
1971. 'Sasanian Bullae
Berlin, 9-10.

GOBL,

GUBAEV, A. GUBAEV,

from the Castle of AJe-depe', Epigrafiya Vostoka {Epigraphy of

the East) 20.

A. LELEKOV, L. 1970. 'Bullae of the Sasanian period from Ak-depe', Karakumskie dremosti (Antiquities from the Karakum) 3, Ashgabat. INVERNIZZI, A. (ed.) 1995. In the Land of the Gryphons, Papers on Central Asian archaeology
in

&

antiquity.

Monographs of Mesopotamia V,
fig.

Florence.

OLIVIER LECOMTE
CNRS,
Paris

ISLAMIDDINOV, M. Kh.
siya), Tashkent, 69,

& SULEIMANOV,

R. Kh. 1984. Erkurgan

(stratigrafiya

i

periodizat-

30.

KOSHELENKO, G.
Boussac

& Invernizzi
A.,
I.

A. 1997. 'Bullae from Gobekly-depe. General Problems and 1997, 377-8.

Main Subjects'.

In

KOSHELENKO, G.
Istory 2, 194-203.

GAIBOV, G.

V.

& BADER, A. N.

1995.

'Two Goddesses?', Vestnik Drevnei

The geographical location of Hyrcania, and more
the border between Iran, the centralized, safe,

specifically

of Dehistan,

at

and

principally sedentary world of the

KRUGLIKOVA,

from Dzhiga-tepe', Drevnyaya Baktriya (Ancient Bactria) 3, [Materials from the Soviet Afghan Archaeological Expedition] Moscow. LUKONIN, V. G. 1971. 'On the subject of Bullae from Ak-depe', Epigrafiya Vostoka (Epigraphy of
T.

1984. 'Bullae

Iranian Plateau, and Turan, the mobile, disturbing and threatening horizon of the steppes, caused it to play a major role in the defence of both the Parthian and Sasanian
empires.

the East) 20.

MASSON, M.
Istory 4.

E.

& PUGACHENKOVA, G. A.

1954. Imprints of Seals

from Nisa', Vestnik Drevnei

region relevant to this presentation consisted in antiquity of a cultural and sometimes administrative entity known in historical periods as Vehrkana or Hyrcania.

The

NIKITIN, A.

B. 1993. 'Parthian Inscriptions

on Bullae from Old

Nisa', Arkheologicheskie vesti

The
its

plain of

Gorgan

constitutes

its

southern part; the plain of Misrian, or Dehistan,
(Fig. 1) are the Elburz in the south, in the

St. Petersburg, 103-6.

northern part.

SEMYONOV,
gemmy
East),

G. L. 1999. 'A Treasure-house in Paikend', Tezisy dokladov konferentsy— Drevnei i kamni Vostoka (Papers delivered at the Conference— Ancient Gems and Stones from the

The geographical borders of Hyrcania

Moscow, 21-2.

north the Great Balkan and the Uzboy, an ancient bed of the Amu Darya, the banks of the Caspian Sea to the west and the Kopet Dagh to the east. The north and south of Hyrcania are very different in appearance. The plain of Gorgan (Fig. 2) in the south

from the proximity of the Elburz, from which flow tributaries of the Gorgan which provides abundant water, added to which is rainfall. It has always been a prosperous agricultural region, easy to cultivate. The northern part of Hyrcania (Fig. 3), on the other hand, is an alluvial plain whose climate is semi-arid. Despite the presence of the Atrek river and its tributary, the Sumbar, in the south-east part of the
benefits
river,
is possible without irrigation. of knowledge, the plain of Gorgan, which possesses more At our present stage favourable hydrological resources and climatic conditions, was occupied earlier than

Misrian plain, no agriculture

the plain of Misrian to the north.

The

latter

was not occupied before the Iron Age,

about fourteenth-thirteenth centuries bc. The plain of Gorgan, on the other hand, was continuously occupied from the Neolithic period (Djeitun type), and very densely during the Bronze Age, from the beginning of the third to the beginning of the second
millennia. This
is

in any case the image provided

by

its

two main

sites,

Shah Tepe and

Proceedings of the British

Academy

133, 295-312.

© The

British

Academy

2007.

328

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

329
in the

Close comparisons for the architectural development of the defences certainly exist nearer to Merv, although few sites have been systematically excavated or published. Ai Khanum in Afghanistan (Bernard 1982; Guillaume 1983; Kritt 1996, 34; Narain 1987; Schlumberger & Bernard 1965: and Lecuyot, above, pp. 155-61) and'Antiochia-in'
Scythia (Bernard 1992, 91-2) certainly merit comparison, and parallels have been drawn between Merv and Bactra (Bernard 1992 and Francfort 1988). In the later phases the development of the Parthian and Sasanian defences

KOSHELENKO,
KRITT,

G. A.,

GUBAEV, A, BADER,

A. N.

& GAIBOV, V

A. 1994. Ancient Merv

testimony of written sources, Ashgabat.

at

of Bactria, Lancaster PA. W. 1979. Aims in Greek fortification, Oxford. LERICHE, P. 1986, Fouilles d'AiKhanoum V, Les remparts et les monuments associes, XXIX, Paris. MCNICOLL, A. W. 1997. Hellenistic Fortifications from the Aegean to the Euphrates; with revisions and additional chapter by N. P. Milner, Oxford. NARAIN, A. K. 1987. 'On the foundation of Ai- Khanum: A Bactrian Greek city', India and the
B. 1996. Selucid coins

LAWRENCE, A.

Gobekly-depe (Koshelenko, above, pp. 269-83) show similarities in the development of the corner bastion and curtain walls with internal corridors divided by transverse walls, while Sasanian walls at Chilburj (Gaibov et ah 1990) and the broad layout of Firuzabad (with a similar street and gate layout) would certainly merit further research.
Overall, our excavations strongly suggest that broader trends in military architecture influenced the development of the city walls of Gyaur Kala. The walls of Erk Kala

Ancient World, Pollet

(ed.),

Leuven, 115-30.

PUSCHNIGG, G. 2006. Ceramics of the Merv Oasis: Recycling the City, London. PUSCHNIGG, G. forthcoming. The ceramics from Merv 1: The Achaemenid to late Sassanian ceramics
from Merv
c.

6th century

bc

to 7th century
P.

CE, London.
1965.

SCHLUMBERGER,
SIDKY, H.
95-118.

D.

& BERNARD,

Ai Khanoum',
to

Bulletin de

Correspondance

Hellenique 89, 590-657.
2000. The Greek
S.

Kingdom of Bactria. From Alexander

Eucratides the Greats Oxford.

and Gyaur Kala shed

on the development of city fortifications in Central Asia over a period of 1,300 years and provide considerable scope for further study.
light
Note.
I should like to express

TASHKHODZHAEV,
USMANOVA,
YuTAKE VYAZIGIN,
260-75.
S.

1963. 'Section through the city wall in Gyaur-kala', Trudy

YuTAKE

12,

Z.

I.

1989. 'The section through the fortress wall in Erk-kala,

Old Merv', Trudy

my gratitude to the British Academy, the organizers of the conference,

19,

21-61.

for giving

me the opportunity to outline some of the results obtained during fieldwork between 1997-2003. This was undertaken under the auspices of the Internationa! Merv Project (1992-2000),

A. 1949. 'The wall of Antiochus Soter around ancient Margiana' Trudy

YuTAKE

1,

Kakamurad Kurbansakhatov, Georgina Herrmann and St John Simpson, and the Ancient Merv Project (2001 onwards), directed by Kakamurad Kurbansakhatov and Tim Williams, both joint Turkmen-British Archaeological Collaborations. For preliminary reports on the annual seasons see Iran 31 onwards. I should like to thank the Iran Heritage Foundation and the Committee for Central and Inner Asia for funding my study visit to London from February to August 2004. I should also like to say how much I appreciated the support of staff of the Institute of Archaeology UCL, and the British Museum, both in the field and in the preparation of the publication.
directed by

BERNARD,

P.

1982.

An

ancient Greek city in Central Asia', Scientific American Ancient Cities

special edition [1994] 5(1), 66-75.

BERNARD, P.
FRANCFORT,
GAIBOV,
21-36.
V.,

1

992. Alexander

Asia', History

and his successors in Central Asia: Part Two, the Seleucids in Central of Civilizations of Central Asia: the development of sedentary and nomadic civilizations,
Persia,

700bc to ad 250, 2, Paris, Unesco, 88-97. H. P. 1988. 'Central Asia and eastern Iran', Cambridge Ancient History: Greece and the western Mediterranean c. 525-479 bc, 4, ch. 3, Cambridge.

KOSHELENKO,

G. A.

& Novikov, S. &

1990. 'Chilburj', Bulletin

of the Asia

Institute

4

GUBAEV,

A.,

KOSHELENKO,

Murghab

Delta, Preliminary Reports 1990-95,

GUILLAUME,

TOSI, M. (eds) 1998. The Archaeological Map of the Rome. 0. 1983. Fouilles d'AiKhanoum, Memoires de la Delegation archeologique francaise en
G. A.

Afghanistan 26, Paris.

326

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA
Development
of*

327

The Phase 2

wall

the defences

It is possible that

Merv's location, close to Parthia, necessitated the repair and modifica-

tion of the fortifications during the Graeco-Bactrian period. In addition, advances in

The

torsion artillery at this time

may have led to the shift from hollow to solid walls, and to an increase in the width of the upper walkway, so that small catapults could be set up on the wall. The Graeco-Bactrian period in Margiana lasted for approximately a century.
There was no clear archaeological dating for
this phase.

revealed a complex sequence of defences, with no fewer almost 1,000 years after than seven walls (Phases 1 to 7), which protected the city for the walls were substantially Alexander. In three phases— Phases 2, 5 and increase the thickredesigned, changing the defensive system: the overall trend was to the width of the wall at its base was 8.80 m. ness of the defences. In Phases 1 and 2 and during Phase 4 the wall During Phase 3 the thickness had increased to c. 13

excavations at

Merv have

6—

m

achieved

its

maximum

width of some

1

5

m.

The Phase 3
This wall

wall

may have been constructed in the early Parthian period, when Margiana became part of the Parthian empire, probably during the reign of Mithradates I (171-139 bc) or Phraates II (138-129 bc). A copper obol of Diodotus I or II (250-239 bc), and a range of pottery, including grey wares, were found in the pakhsa blocks from
the wall: but these were probably residual, as the blocks contained large quantities of
cultural debris.

The Phase 1 walls are striking, especially as they represent a virtually intact, threefollowed ideas of milistorey defensive wall. The construction appears to have closely
tary architecture recorded

by Philo of Byzantium,
is

century bc (Lawrence 1979, 71). This
settlers in the

second half of the third hardly surprising, given the quantity of Greek
in the

region and

the role of the

Greek hierarchy

in the region (Sidky 2000,

130-2).

had a Greek inscription, the letters manner resembling that used in Egyptian papyri of the of which were written in a second-first centuries bc (Sergei Remirovich Tokhtasiev, pers. comm.).
brick found in the brickwork

A mud

brick walls with the stone walls of Hellenistic cities in striking similariAsia Minor, such as Perge, Side and particularly Caunus, also reveals of the embrasures and suggested merlons. At ties, especially the dimensions and shape structural features, together with information Caunus, the combination of tactical and

Comparison of

these

mud

drawn from other

The Phase 4

wall

The Phase 4
the late
first

wall

may

also date to the Parthian period.
first

Two Parthian-Margiana coins of
bricks

century bc or the
its

century

ad were found between
was
at that

A. W. McNicoll to date the defences to the middle or the end of the fourth century BC (McNicoll 1997). second storey: for examParallels can also be found for the arched chambers in the have been dated by McNicoll to which ple, the stone defensive wails at Perge and Side,
sources, led

and provide a

pointer to the date of

construction. It

time that Margiana was endeav-

a date no

ouring to achieve autonomy from the Parthian Empire, which could have provided the

impetus for the strengthening of the defensive wall. Another possible political impetus

might have been the
seeking to expand.

state

of relations with the Kushan empire, then

in the

ascendant and

was than -225 bc (McNicoll 1997). The use of arches in curtain walls Rhodian walls of his day. also noted by Philo of Byzantium, who used as a model the also points out that the Rhodian walls were McNicoll, when mentioning that fact, post quern for the destroyed by the earthquake of 227 bc, suggesting that as a terminus married to arches' which Philo describes (McNicoll 1997). These paral'curtains
earlier
. . .

lels

The Phases 5 and 6

walls

century date would seem to concur with the written during the reign of sources and suggest that the first wall at Gyaur Kala was built govern the Upper Satrapies of the perhaps soon after he began to

of a

late fourth or third

Antiochus

I,

These two rebuilds of the Phase 4 wail may have occurred during the early and middle Sasanian periods. Five coins, minted during the reign of Shapur II (310-379), and one coin of Bahram IV (388-389) or Kavad I (488-531), were found in various contexts in
the middle gallery.

Seleucid state in

293 bc.
fortifications in the Hellenistic world, however, reveals that

Comparison with stone

Phase 7 wall

major differences, probawhile there are similar architectural elements, there are also For instance, the base of bly resulting from the different building materials employed. mud brick walls is usually battered, while stone walls have more vertical sides; the use brick walls; the of wide platforms, extending some distance beyond the line of mud form of steps; the different possibilities for interior finish of the base of the wall, in the
cheapness of modification of the walls, stemming from the availability and relative cycles of earthen earthen building materials; coupled with the different maintenance

A coin of Kavad I (499-531) was found during YuTAKE excavations of the Phase 7
wall.

and stone constructions.

324

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR
Phase 10

KALA

325

the proteichisma, while another conbetween the bastion and the proteichisma. It is probable that there was an upper walkway with parapet, above the upper gallery, which was destroyed during subsequent rebuilding.
trolled the space

One arrow-slit was angled to cover the doorway in

Probably after the walls had ceased to function as a defensive circuit, a number of burirecorded als took place along the walls. One burial, on the south side of the city, was burials were also found on the west wall by the by the IMP, and some eighth century

YuTAKE team (Tashkhodzhaev
Phase 5

1963).

During

front (west) of the wall

was once again radically reorganized. The platform to the was modified by the truncation of the Phase A proteichisma and the construction of an earth slope, which blocked the lower gallery arrow-slits of the Phase 4 wall. (The lower gallery, having ceased to function as a defensive feature, was
this phase, the wall

reused for the deposition of excarnated burials in accordance with the Zoroastrian rite.) The top of the slope lay immediately below the Phase 4 upper gallery arrow-slits,
suggesting that Phase 4 may have continued to function. Any modifications upper part of the curtain wall were lost due to later rebuilding.
to the

The sequence of defensive walls can be considered against a background of historical events, while new archaeological dating evidence, both numismatic and ceramic, can be
used to assess the overall development of the sequence. There are two key historical events that are relevant to the defensive circuit of Gyaur Kala. The first is the accession of Antiochus Soter, as joint ruler in 293 bc and then as emperor from 281-261 bc, and his purported construction of substantial
fortifications at

Phase 6
phase the gallery of the Phase 5 wall was substantially internally remodelled, with the insertion of a low vaulted roof. This blocked the arrow-slits, which were
this
infilled,

2000, 133).

Merv (Pliny the Elder Naturalis Historiae VI, 18, 46-7— see also Sidky The second was the capture of Merv by the Arabs, in 651 ad.

During

but presumably supported the reconstruction of the gallery above. The old

gallery

m high, to provide access to the corner bastion. The floors in the corridor were resurfaced with earth on several occasions: the latest floor was m higher than
and
1.90
1

now appears simply to have functioned as an access corridor, some 0.90 m wide

Numismatic data can provide useful dating evidence, although such data needs to be used cautiously. There was considerable reuse of cultural debris at urban sites, especially for building material, and there is a strong possibility of residuality (Puschnigg 2006). Eighty-one coins were identified by N. Smirnova from this trench but only 10
were found in securely stratified contexts, relevant to the dating of the different phases. The pottery has been processed and studied by Gabriele Puschnigg, who has examined the typological development in accordance with the phases described above.
Diagnostic shapes from a wide range of sequences provide a relative dating sequence, with some suggestions of absolute dating (Puschnigg forthcoming).

the original floor, by which time

it

would have been possible only to crawl along the
to ventilation.

passageway, or perhaps

it

might have been limited

Phases 7 and 8

Two

further phases were identified on the basis of results of the YuTAKE excavations (Tashkhodzhaev 1963). However, the state of preservation of the Phase 7 wall is poor, and it was not possible to confirm the published YuTAKE information. The Phase 7
wall

In addition, the stratigraphy of the defensive sequence was not always straightforward, particularly the infilling of chambers and passages, the refacing of walls, and the stepped interior platforms. These often possess sequences of their own, not
directly linked to the
wall.

main

circuit walls or the outer platforms, except for the

Phase

1

was apparently erected on a low platform, which extended
been
rebuilt (Phase 8).

to the west

and seems

to have

The Phase
It is

1 wall

suggested that this wall dates to the Seleucid period, based on the written sources and
in defensive installations in

Phase 9

Our excavations behind
for a

the bastion established the presence of a corridor, belonging to

the Phase 7 wall, filled with

mud

bricks.

new wall, Phase
late walls

9.

However,

this wall

This action was presumably taken to prepare has either not survived or was never built.
side of the city.

Mesopotamia and Asia Minor (see below). The ceramics from this phase belong to the Yaz HI period, which Puschnigg broadly considers supportive of that date. If Margiana was part of the Graeco-Bactrian state, then the Seleucid period at Antiochia Margiana might have lasted
supported by architectural parallels found
until
c.

Remains of

do survive on the north

250

bc.

322

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

323

Figure 8.

Protekhisma in front of the Phase 4 wall (looking west).

Phase 4 (Figs 8 and

9)

was extended and raised, to carry an outworks or protekhisma, the curtain wall was substantially remodelled and the corner bastion was concomitantly extended.

Tn Phase 4 the western platform

To
thick,

the west of the Phase 3 wall a new external platform was constructed, c. 2 supported on massive pakhsa blocks, each 0.75 high, laid on courses of mud

m
Figure 9.

m

bricks 0.60-0.80

m high.

The platform was

built

up

against the Phase 3 wall

and the

semicircular corner bastion.

Reconstruction of the Phase 4 wall.

Aproteichisma surviving to a height of c. 0.45 above the platform, complete with and an outer walkway 1.5 wide, was built in front of the corner bastion and curtain. There were seven surviving arrow-slits, spaced at varying intervals and
arrow-slits

m

m

must have been via the upper
into the corner bastion.

gallery, in

which an arched doorway was found leading

The arrow-slits sloped downwards and were designed for archers, who could strike the enemy not only at long distance but also if they managed to penetrate the outer walkway. At the bend in the
fire.

slanted in different directions to cover a wide field of

There were rectangular
of the lower
gallery.

arrow-slits, 0.40

The bottom of

the

slits

m wide and 0.80 m high in the outer wall was 2.0 m above the level of the platform,
more than 2.0 m. A bastion, had a pointed

direction of the corner bastion a

doorway had been incorporated into the protekhisma, which made it possible for the defenders to clamber out to the edge of the platform or hide behind the protekhisma. The outer edge of the platform was destroyed, probably
was a of vaulted chambers in the lower gallery, the arrangement of which is reminiscent of the vaulted chambers of the Phase 1 walls. Although they were higher and larger, they must also have been linked by doorways. Part of the chamber in front of
series

suggesting that the height of the protekhisma would not have been
false arrow-slit,

top formed by two sloping bricks: the top of the false
the real

when the canal was dug at the beginning of the twentieth century. The curtain wall consisted of two galleries, an upper and a lower

found immediately next to the junction with the slit was 1 m higher than that of arrow-slit, suggesting that the real and false arrow-slits may have formed a
gallery

one. There

chequerboard pattern.

The upper

was 2.75

m wide and had
pillars

a

rammed

earth floor.

The

gallery

may

have consisted of sections separated by
defenders to

along the outer wall, one of which has been
they penetrated the protekhisma.

found. Five arrow-slits were found in the outer wall and were arranged to enable the
fire at

the bastion was excavated, although

no doorway was found, and

access to the bastion

enemies over a long distance and

if

320

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

321

Figure 6.

Reconstruction of the wall

in

Phase

2.

The stepped
another 4.5

inner base of the Phase
into the city

1

wall

was

also modified.

A new step extended

m eastward

platform, raising the ground by 0.50 m.
their original form.

and covered the two lowest steps of the original The three upper steps may have continued in
1

from its rebuilt parapet and the raised external platform, would appear not to have been substantially modified, the Phase 2 wall had been transformed into a solid wall and was about 1 m
wall, therefore, apart

While the external appearance of the Phase

Figure 7.

Reconstruction of the Phase 3 wall.

higher than

its

predecessor.

Phase 3

(Fig. 7)

A new platform, built of mud bricks laid on top of two rows of pakhsa blocks, was constructed against the exterior face of the Phase 2 wall.

The platform was

1.25

m thick

and gradually sloped down to the west (beyond the edge of the section excavated in 2003: which leaves questions regarding its full extent and how it relates to the moat). Within the platform the edge of some brickwork was uncovered. It stood 0.75 m high, turning southward. The gradual curve of the brickwork suggests that it was semicircular: perhaps it was the foundation of a corner bastion.

were refaced with pakhsa blocks. The bottom row extended some 3 m from the earlier wall face, while the upper rows gradually decreased in width to form a slope, which ran roughly halfway up the inner face of the wall. The at the top. The bottom of the at the base and 4 new width of the wall was c. 13 finished with three steps, the bottom one interior surface, as in previous periods, was

The

interiors of earlier walls

m

m

being

c.

5.5

m wide.
was considerably
larger.

probable that a new parapet was built at the top of the wall but was truncated the construction of the next wall. The Phase 3 wall was similar in construction during
It is

to the Phase 2 wall but

318

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA
and
this

319

combination of embrasures and buttresses suggests that the parapet would

have been finished with merlons. The upper walkway was supported by the vaulted

mud

brick ceilings of chambers

forming a middle floor within the wall. Three chambers were cleared in the exposed wide, c. 2.30 m high and extended into the wall for a section: these were 1.40-1.50 wide, behind the vaulted chambers, depth of 1 .50 m. There was a walkway, 1 10-1 .20

m

.

m

enabling defenders to make
ably for use with slings

their

way along the

wall.

A large number of projectiles were

found on the floor of one of these chambers: the projectiles were of various sizes, proband catapults. The absence of arrow-slits in the middle floor
suggests that these chambers were used for storage

and

that defenders

would have

fought from the walkway along the top of the curtain wall. The base of the Phase 1 wall within the city was constructed to form five or six steps

ascending over a distance of 4.5 m: the two upper steps were 0.75
0.30

m wide and 0.25 and

m high. Different clay had been used for their construction to that of the main wall

foundation or the outer platform. The stepped base of the wall could have facilitated
Figure 4.

The walkway, parapet and unexcavated arched chambers of the Phase

1

wall (looking west).

movement along the wall, or possibly acted as a staircase, providing access to the upper
wall.

Phase 2 (Figs 5 and 6)

During the second phase the defensive wall was radically rebuilt. The Phase 1 chambers of the middle storey, and the adjacent walkway, were filled with clay and mud bricks up to the level of the floor of the upper walkway. This was probably intended to facilitate the construction of a wider base (1.75 m wide) for a new upper walkway, which was raised by c. 0.30 m. Between the straw-and-clay replastering of the walkway floor a collection of unfired slingshots was found. The Phase 1 parapet, south of the south buttress, was truncated and its height raised with two rows of bricks, so that it measabove the floor of the walkway of the Phase 2 wall. ured 0.80 wide and was 2.40

m

m

The

original

Phase

1

parapet to the north of the rebuilt section, however, continued to

be used and was simply replastered.

While the western part of the outer platform of the Phase 1 wall continued unchanged the eastern half was raised by 2 m against the base of the wall. Tashkhodzhaev considered that this was simply designed to make it harder to dig under the wall (1963): although possible, this is unlikely to have been the sole reason for this
Figure 5.

The

inner face of the parapet of the Phase 2 wall (looking west).

modification.
attacker

The main wall was primarily built of mud bricks and stood c. 10 m above the platform. It had a walkway, 1.10 m wide, with a parapet 1.75-2.0 m high, consisting of

More likely would have been the creation of yet another obstacle for the by creating a steep additional climb of approximately 1.5 m from the surface of the Phase 1 platform. The addition to the platform not only made access to the wall more difficult but also enabled the city's defenders to change the angle from which they
fired while

m apart (Fig. 4). An embrasure, surviving near the north buttress, was 0.80 m wide and 1.20 m above the walkway
courses of single

mud

bricks, strengthened

by buttresses 2

standing on the upper walkway.

It

should be recalled that the width of the

walkway
used.

at this stage

had been

increased, perhaps so that personal artillery could be

floor.

Another embrasure would probably have been existed south of the south

buttress,

316
projectiles; nevertheless, the
I

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

317

Period
I.

II wall

was assigned by Usmanova to Antiochus

(281-261 bc), son of Seleucus

Usmanova dated

the Period III wall,

and a substantial tower built against the outer

edge of the Hellenistic wall, to the Parthian period. The tower was made of square mud thick). The base of the bricks, with two thick layers of pakhsa (in places more than 1

m

tower rested on a layer of sand above the berm

of the

Achaemenid wall. The

fact that

the tower was built against the Achaemenid wall, the lack of dating material from the Parthian period, and a parallel with the monolithic towers at Ai Khanum (Leriche 1986, 88-91), however, perhaps suggests that the tower should be dated at least to the

beginning of the Seleucid period, or to when the Greeks arrived
the last quarter of the fourth century bc. If such

in

Margiana, that
is

is

to

an interpretation

correct, then the

report by Pliny the Elder concerning the building activity of Alexander, might acquire

concrete archaeological confirmation; although

it is

also possible that the city destroyed

by the barbarians might have been
the building of the
first

by Antiochus Soter, perhaps making defensive walls of Gyaur Kala, and the rebuilding of the walls
re-established

of Erk Kala, simultaneous.

Gyaur Kala
In 1957-1958 the
first

archaeological study of the fortifications in the south-west cor-

ner of Gyaur Kala was undertaken by a
(1963).

YuTAKE

team headed by
at the

S.

Tashkhodzhaev
Figure
3.

A section was cut
site for

through the walls where,

beginning of the twentieth

Reconstruction of the Phase

1

wall.

Batmanov had breached the wall to bring water into the A small trench was dug at the base of the earliest wall, conclude that the foundations had been built over alterwhich led the researchers to nating layers of tuyun (a local name for natural clayey deposits) and sand, formed by the old channel of the Murghab river. Of interest is Tashkhodzhaev's conclusion that water would have come into the city via a canal at this point, and that the east bank of the canal formed the western edge of the pakhsa platform of the Seleucid wall, and thus could be assumed to be of the same date. The YuTAKE excavations did not go further
century, a technician called

ancient city

farming.

author in 2002 and 2003.
material

New data has made it possible to review the earlier YuTAKE
new phasing structure for
the development of the city wall.

and

to propose a

The development
Phase
1

of the city walls of

Gyaur Kala

(Figs 3

and 4)
less

than the bed of the canal.

During the excavations four periods of wall were identified. The earliest was associated with Antiochus I, while the second wall was assumed to date from the early Parthian period, to the reign of Mithradates II (123-87 bc). The third wall was dated to the late Parthian period, while during the reign of Shapur I (241-272 ad) this wall was built higher (not given a period). The fourth and last wall, according to YuTAKE, was dated to the sixth century ad, to the reign of Khusrau I (531-579 ad). Two stages
were
identified in the building

west was constructed.

easthigh and extending some 27.5 to 28.0 than 3.50 upper surface sloped at the western edge, from the base of the subsequent wall to a north-south moat. The platform consisted of thin layers of brown clay, brown clay mixed with sand, and thin sandy layers, and was perhaps quarInitially

a platform, no

m

m

Its

ried

from the cutting of the moat. The platform under the wall was

built

by horizontal

of

this particular wall.

From 1997-2003

the International

Merv

Project re-examined Tashkhodzhaev's sec-

between which thin layers of almost pure brown clay were deposited, and demonstrate that different construction techniques and materials were used for the foundation as opposed to the outer platform. The horizontal nature of these
layers of sand,

tion through the fortifications

(MGK

6) (Fig. 2).

These excavations were jointly

alternating deposits had been interpreted

by the

YuTAKE

team as deposits from the

directed by St John Simpson and the author (1997 to 2000) and continued by the

Murghab (Tashkhodzhaev

1963).

314
the north wall (Fig.
enclose an area of

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

315

1). Gyaur Kala's fortifications measure nearly 8 km in length, some 340 ha and still stand some 1 8 m high. Mounds along the walls

Background
Evidence from Graeco-Roman written sources relating to Merv
his History
is

indicate the remains of defensive towers, perhaps as
five gates, located in the centre

many as eighty. There were at least
city,

meagre, and

it is

not

of each side of the

except in the north where the

possible to put together a satisfactory picture of the political history of the region. In

gate was located to the north-east of the citadel,

Erk Kala. Roads divided the city into four quarters, and a basic rectilinear Hellenistic street pattern can still be seen in aerial photographs.
This paper discusses the excavation of a section through the defences in the southwest corner of the city of Gyaur Kala (Fig. 2). Phases of construction and rebuilding
are summarized, along with preliminary evidence for dating individual phases. Earlier

work on Erk Kala and Gyaur Kala,
is

carried out by

YuTAKE between

1960 and 1990,

of Alexander of Macedon, Quintus Curtius Rufus reported that after crossing the Okh and Oxus Rivers (identified as the Tedzhen and Amu Darya), Alexander came to Margiana, where he founded six fortresses on high mounds. In Strabo's Geographia mention is made of a region called Margiana in a plain surrounded by deserts, which was visited by Antiochus Soter: Amazed by the fertility of the plain he gave orders that it should be girt about with a wall which would measure 1,500 stadia and he founded the city of Antiochia' (Strabo XI, 10, 2). In Pliny the Elder's Naturalis

also summarized.

Margiana stated: 'encircled by a magnificent and with a circumference of 1,500 stadia. Here Alexander founded the city of Alexandria, which was destroyed by the barbarians, but Antiochus, son of Seleucus, established a Syrian city on exactly the same spot, which is crossed by the River Marg, which flows into Lake Zotkhus and he preferred to give it the name Antiochia. Its circumference measured exactly 70 stadia' (Koshelenko et al. 1994).
Historiae (VI, 18, 46-7) a description of
ring of mountains

Despite the valuable character of these reports their interpretation

is

not straight-

tM
fyi&fe^Vj

forward. Archaeological evidence confirms the presence of Greeks in Margiana, and the

;QJ

m

m

by teams of the South Turkmenistan Archaeological Multi-disciplinary Expedition of the Turkmenian Academy of Sciences (YuTAKE: Vyazigin 1949), and recently by an Italian -Turkmen project (Gubaev et al 1998). These broadly support the reports provided by ancient
studied

Merv oasis was indeed surrounded with a wall, which was first

fQi

authors, although the oasis

is

certainly not encircled by mountains.

mm
Previous excavations
Erk Kala

mm

Between 1963 and 1979 a substantial section, 2 wide and 75 long at the base (90 m long if eroded material is taken into account) was cut through the defensive walls of Erk Kala. The excavator, Z. I. Usmanova, identified four main periods

m

m

Figure

2.

South-west corner defenses wall section in 2000 (looking south).

IA and IVA) (Usmanova 1989). The Hellenistic period (II) was represented by apakhsa wall 2.5 m high and 13 m wide, erected on top of the Achaemenid Period I wall to a height of 18.0 m above the contemporary ground surface. Usmanova does not rule out the possibility that the upper part of the wall had been fashioned as a parapet. The pakhsa of the Hellenistic wall differed from that of the Achaemenid wall, perhaps because the earth from which it had been built was not quarried from the moat. The only materials found in the wall were 'nondescript' fragments of pottery and round clay
(I-IV) in the erection of the walls, with two major rebuilds (designated

314
the north wall (Fig.

V.

A. Zavyalov

FORTIFICATIONS OF THE CITY OF GYAUR KALA

315

Gyaur Kala's fortifications measure nearly 8 km in length, enclose an area of some 340 ha and still stand some 18 m high. Mounds along the walls indicate the remains of defensive towers, perhaps as many as eighty. There were at least
1).

Background
Evidence from Graeco-Roman written sources relating to Merv
possible to put together
his History
is

meagre, and

it is

not

five gates, located in the centre

of each side of the

city,

except in the north where the

a

satisfactory picture of the political history of the region. In

gate was located to the north-east of the citadel, Erk Kala. Roads divided the city into four quarters, and a basic rectilinear Hellenistic street pattern can
still

be seen in aerial

photographs.

This paper discusses the excavation of a section through the defences in the southwest corner of the city of Gyaur Kala (Fig.
2).

Phases of construction and rebuilding

are summarized, along with preliminary evidence for dating individual phases. Earlier

of Alexander of Macedon, Quintus Curtius Rufus reported that after crossing the Okh and Oxus Rivers (identified as the Tedzhen and Amu Darya), Alexander came to Margiana, where he founded six fortresses on high mounds. In Strabo's Geographia mention is made of a region called Margiana in a plain surrounded by deserts, which was visited by Antiochus Soter: 'Amazed by the fertility of the plain he gave orders that it should be girt about with a wall which would measure 1,500 stadia

work on Erk Kala and Gyaur Kala, carried out by
is

YuTAKE

between 1960 and 1990,

also summarized.

and he founded the city of Antiochia' (Strabo XI, 10, 2). In Pliny the Elder's Naturalis Historiae (VI, 18, 46-7) a description of Margiana stated: 'encircled by a magnificent ring of mountains and with a circumference of 1 ,500 stadia. Here Alexander founded the city of Alexandria, which was destroyed by the barbarians, but Antiochus, son of Seleucus, established a Syrian city on exactly the same spot, which is crossed by the River Marg, which flows into Lake Zotkhus and he preferred to give it the name Antiochia. Its circumference measured exactly 70 stadia' (Koshelenko et al. 1994).
Despite the valuable character of these reports their interpretation
is

not

straight-

im
:;.;:.i|^^#lSfr 1T^?T'-^'^
.

WjS^^S^^ jflp^||

urn

forward. Archaeological evidence confirms the presence of Greeks in Margiana, and the

Merv

oasis

was indeed surrounded with a

wall,

which was

first

studied by teams of the

„-Vt

f.'*

-

-

1/-

1

;.";.

South Turkmenistan Archaeological Multi-disciplinary Expedition of the Turkmenian Academy of Sciences (YuTAKE: Vyazigin 1949), and recently by an Italian-Turkmen
project

61

(Gubaev

et al 1998).
is

These broadly support the reports provided by ancient
certainly not encircled

*W£|S
"-

authors, although the oasis

by mountains.

K.

:=

=

-

"

JSliy
Previous excavations

-^""c^'-^;^!
Erk Kala

illiS^^^^^H
."-. :

^^^p^^^i^^^^^^^^^^^^^§

Between 1963 and 1979 a substantial
(90

section, 2

m wide and 75 m long at the base
identified four

:

V

-

.

M^,y:
^^^^L^-ikil
%\~r
-[
"

m

long

if

eroded material

is

taken into account) was cut through the defensive
I.

'

J&-

'

'

'

walls of

Erk Kala. The excavator, Z.

Usmanova,

main
I

periods

(I-IV) in the erection of the walls, with two major rebuilds (designated

-^:y£&&%-

(Usmanova high and 13
18.0
sibility that

A and IVA) represented by apakhsa wall 2.5 m 1989). The Hellenistic period (II) was

Figure 2.

South-west corner defenses wall section in 2000 (looking south).

m wide, erected on top of the Achaemenid Period I wall to a height of m above the contemporary ground surface. Usmanova does not rule out the posthe upper part of the wall

had been fashioned as a parapet. The pakhsa / of the Hellenistic wall differed from that of the Achaemenid wall, perhaps because the earth from which it had been built was not quarried from the moat. The only materials found in the wall were 'nondescript' fragments of pottery and round clay

312
could prevent the

Olivier

Lecomte

fall of a state which, like many previous centralized powers, had already started to disintegrate because of internal pressures and dynastic problems. To

these causes should be
controlled.

added the

fragility

caused by the immensity of the

territories

References

ARRIAN.

Arrianus fragments scriptorum de rebus Alexandri

M.

Arriani Anabasis et Indica, edited

by Fr. Diibner. No date. Firmin Didot, Paris, 83. ATAGARRYEV, A. E. 1994. 'Kerptitchlining syry ajanboljar'. Turkmen Medenieti 2, 6-8. BOUCHARLAT, R. & LECOMTE, O. 1987. Fouilles de Tureng Tepe. Sous la direction de Jean
Deshayes. 1 Les Periodes Sassanides et Islamiques,

V.

A.

ZAVYALOV
St.

Institute for the History of Material Culture,

Petersburg

ERC,

Paris.

Empire Perse, Fayard, Paris. CAB IX. 1966. The Roman Republic 133-44 bc, 575. CHRISTENSEN, A. 1971. LTran sous les Sasanides, 1944, reed., Osnabruck. FRANCFORT, H.-P 1988. 'Central Asia and Eastern Iran'. In J. Boardman et al. (eds), The Cambridge Ancient History, vol. 4, Cambridge (2nd edition), 165-93. GYSELEN, R. 1987. 'Bulles et sceau sassanides'. In Boucharlat & Lecomte 1987, 187-91.
1996. Histoire de
I'

BRIANT, P.

Introduction

The historic urban centre of Merv is located in the Murghab oasis in modern day Turkmenistan. During the Achaememd period the first city (Erk Kala) was founded
within a polygonal enclosure wall, which
still

HANSMAN, HARMATTA,

J.

1968. 'The problem of Qumis',
J.

JRAS,

1 3

1-39.

stands to a height of

1996. 'The walls of Alexander the Great
10, 79-84.

and the Limes Sasanicus',

Bulletin

of the

occupies an area of

some 12 ha

(Fig.

1).

A massive city,

more than 20 m and Antiochia Margiana (Gyaur

Asia Institute

HERZFELD, E.
I,

Kala), was built in the Seleucid period; Erk Kala formed the citadel in the middle of

1924. Paikuli

monument and

inscriptions

of the early history of the Sasanian empire

Berlin.

HUFF, D.

1981. 'Zur Datierung des Alexanderwah", Iranica Antiqua

London, 1914, 1976. 9. JUSTI, F. 1963. Iranisches Namenbuch, Marburg, reprinted 1963. KIANI, M. Y 1982. Parthian Sites in Hyrcania. The Gurgan plain,
(ed.).

Parthian Stations.

W H. ScholT

XVI, 125-39.

Isidore ofCharax,

Berlin.

KHLOPIN, I. N. 1983. Istoricheskaia geografiia iuzhnykh oblastrei Srednei Azy, Ashkabad. PAULY WISSOWA 1914. Paulys Realencyclopadie der Klassischen AltertumsWissenschaft, Neue Bearbeitung begonnen von Georg WISSOWA. A. Druckenmuller. Stuttgart.
POLYBIUS.
1960.
classical library,

Anabase d'Alexandre, with an English translation by W. R. Paton, London, Harvard University Press, Cambridge, 1960, 175-9.
1961. Histoires, Texte etabli et traduit par

vol. IV,

Loeb

QUINTUS CURTIUS.
Lettres, Paris, 1, 178.

H. Bardon, 2nd

ed.,

Les Belles

1. The cities of Merv. The Achaememd city of Erk Kala became a citadel of the Hellenistic of Antiochia Margiana (now known as Gyaur Kala). The later Islamic city of Sultan Kala developed to the west. The location of the excavation of the defences of Gyaur Kala, the subject of this

Figure
city

paper,

is

shown

inthe south-west of the

city.

Proceedings of the British

Academy

133, 313-329.

© The British Academy 2007.

Sign up to vote on this title
UsefulNot useful