The Circle of Ancient Iranian Studies
IRANIAN ARCHITECTURE: ARSACID DYNASTY
Parthian Monuments of the Greater-Iran
(Transcaucasia and Central Asi)
By: Burchard Brentjes
problem of "Parthian" art is fairly often argued in relation to
monuments found in mainland-Iran and Mesopotamia dated in the period of Parthian
dynastic domination, while several important sites in northwest and northeast of
the Greater Iran, i.e. in Central Asia and Transcaucasia have received much less
attention if mentioned at all. The language barrier to using Soviet publications
has meant that excavations made even decades ago have not received the attention
they deserve. Here I limit myself to two temple sites and one residential, those
of Mansur depe, Dedoplis Mindori, and Nisa, although others are mentioned in
connection with them.
The two temples of Mansur-depe and Dedoplis Mindori are open air monuments; that is, the largest element of each is an unroofed court. No similar structure has been found, at least until now, at Nisa. But Mansur-depe is sufficiently near that site to be taken as a central sacred place in connection with the residence of Parthian kings at Parthava-Nisa, known as Mithridatkert, and we will see several trends in its architecture which connect it with Hatra and Ctesiphon. Koshelenko dates the Mansur-depe temple on the basis of the ceramic material to the last centuries BCE although it may have continued in use until the end of the Parthian dynasty.
depe (Mansūr Tepe)
is located near the road from Ashkhabad to Firuz, about five kilometers from
Nisa. A court 175 meters from west to east and 160 meters from north to south
(fig. 1) was found in the center of the town (of undetermined size; the site was
only tested. The enclosure, measuring 190 x 170 meters, is much smaller than
that at Hatra (456 x 320 m/, but the similarity is beyond doubt.
court at Hatra first was thought to be a caravanserai,
but examinations of Pasargadae and Bard-i Neshandeh
support the utilization of the courts for Iranian cult practices. Their
importance may be seen as similar to that of the Kafir "dancing
places" of today.
Hatra and Mansur-depe contain large ivan (ayvān) style buildings,
at the sides of the courts, which Andrae took to be palaces. But the excavations
of the Department of Antiquities of Iraq at Hatra since 1951 have proved that
they were shrines, part of the temple of the sun god.
The same must be true for the structure at Mansur-depe. As at Hatra, ivans are
the dominant architectural features; their basic ritual character seems to be
The organization of space is different from that at Hatra. Instead of Greco-Roman temples, the buildings are of the eastern type with circumambulatory corridors (fig. 2). Those at Mansur-depe were built into the corners. The northwestern building is approximately 7.5 meters square. The southwestern structure was not excavated. Uncovered in the northeastern corner was an ivan with abutting corridors, 7 meters square. The most interesting building was found at the southwestern side. It was dominated by a large ivan that opened on the court. The plan of this main structure at Mansur-depe can be compared favorably with the Taq-i Kisra at Ctesiphon, rather than with the Hatra "palace." The main rooms of both are more or less similar, an ivan before a rectangular, walled hall. They differ in size: that at Ctesiphon measuring 49.92 x 26.25 meters, at Mansur-depe 9.50 x 8 meters.
of wall painting and stuccowork were found at Mansur-depe, but they have not
Schlumberger identified Hatra with the city of Shamash, the Semitic form of
Mithra, the sun god. What of Mansur-depe? It could have been the site for the
Mithras fire. Was it, then, Mithridatkert? The equating of the southwestern
building at Mansur-depe with the Sasanian palace at Ctesiphon might be taken as
an argument against the ritual interpretation of the Hatra structure,
particularly since otherwise we do not have a palace at Hatra.
Mansur-depe "palace" provides the Parthian prototype for Ctesiphon.
The royal residence of the Sasanians had a ritual character as the residence of
the "ruler of the world" whose throne stands on the "giant" Bahāristān
carpet filling the floor as a "garden;" the symbol of the world.
Parthian kings may have been identified, or at least connected, with Mithras, as
shown by the many theophoric names. Mithras seems to have been the god of the
dynasty, and, hence, Mansur-depe may have been the location of the Parthian
dynastic fire of Mithras, much like the Shamash temple at Hatra.
open air temple complex of Parthian times was found in eastern Georgia at
Dedoplis Mindori (fig. 3/).
It stands in the valley of the Pron River, a tributary of the Kura. The complex
covers an area 255 x 135 meters, the long side on the north-south axis. An open
court, 104 meters square, occupied the central section of the complex. Three
buildings were grouped on the northern side. In the center was an ivan temple,
and three similar halls stood in the northeastern corner. The main temple seems
to have been on the southern side, a brick structure 46 x 30 meters. The cella
was square, 17.2 meters in area, the promos 17.2 x 11 meters. Four pillars,
perhaps of wood on post-Achaemenid bases (fig. 4), carried the roof of the
portico. On the western side, as on the eastern, was a temple of identical plan:
an open hall with four columns and a hall with two columns on the court side.
There is a (non-correlated) carbon-14 date of 115 BCE from an unidentified
Iranian character of this temple is indisputable, while its size indicates the
importance of the site which may have been a northwestern center of the Parthian
empire. Parthian influence is found even further to the west. The graves of the
"Pitiaxes of Armasi" (the Castle of Ahura-Mazda) in central Georgia
contained gems and silver work showing Parthian influence, such as the
Aspauroukis gem (fig. 11) and the plates with horses standing in front of an
Large areas serving as ritual centers are known from southern Choresmia, the low mountains of Kiiseli-Gyr and Kalaly-Gyr dated to the sixth and fifth centuries BCE; the first square, the other triangular. Both must have been ritual centers of the early Saka federation.
connections with the northern steppes and Choresmia can be found in the
architecture of Parthava-Nisa, a proto-Parthian settlement that may have been
the center of the Achaemenid satrapy of Parthava. After the Parthian occupation,
a royal palace (now known as Old Nisa) was built beside the settlement New Nisa)
/fig. 5/. Old Nisa stands on a platform about 7 meters high (fig. 6/. The
fortification wall had five corners and was set with towers. The northeastern
side measured 350 meters, the northern side 100 meters, the southwestern side
540 meters, and the southeast side over 250 meters. A massive bastion was
erected in the southeast corner. Two ramps led up to the entrance of the
good deal has been written about the buildings of the citadel, but many problems
still remain (figs. 7, 8/. A long building at the southeastern wall is thought
to have served as a barrack. In the northern part stood a quadrangular house,
59.7 meters on a side. Twelve chambers enclosed a 38 meter square courtyard with
peristyle. The chambers, of equal size, had clay benches along the walls and
four columns supporting the roof. The building has been identified as a
palaestra, a treasury, or as a mortuary palace because of the richness of the
finds: no less than forty ivory rhyta. The rhyta carry carved decoration that
combines Greek, Iranian, and nomadic elements. They may have been made in
Bactria rather than in Parthia.
main complex at Nisa is only partly excavated, but the rooms and buildings are
dissimilar to those we know from Parthian sites in Iran and Iraq. A
representative hall is 20 meters square. Three doors open onto a richly
decorated two-story room with half-columns. Ceramic tiles decorated the hall.
Next to this chamber, thought to be a memorial hall for the ancestors of the
dynasty, stood a strange tower. The center was made up of a massive, square clay
socle, 20 meters long and at least 6 meters high. A corridor surrounded the
socle and two towers stood at the facade. At the western side was another
strange square structure, 25 meters in length, enclosing a round hall with a
cupola 17 meters in diameter. Two entrances nearby were "hidden." The
room was adorned with clay statues and tiles. North of the main hall stood an
open hall with two rows of four columns, reminiscent of the Achaemenid apadana.
A wine depot was found next to the "square house." In the settlement
itself, a rectangular building was uncovered next to the town wall.
elements appear in the architectural decoration and the peristyle as well as in
many of the works of art, such as the Greek statues (fig. 9).But the
architecture bears no relation to the Hellenistic tradition. The round hall was
thought to be derived from the round temple of the Greeks, but the Hellenistic
temple was built to be seen from the outside; here, the round room was concealed
within a clay block, and surrounded by a corridor for the circumambulations of
the cult processions, a concept alien to the Greeks. The prototype may have been
the Asian yurt, ritualized in the steppes in the early Saka tombs of Tagisken
(eighth-seventh centuries BCE) and the late one at Balandy II (second century BCE
) in Transoxiana. The most representative building in this tradition is the
temple-tomb of Koi-Krylgan-Kala in Choresmia, similar to the round house at
Shahr-i Qumis, a ritual house of a city thought to be Hekatompylos.
Returning to the ruins at Nisa, we might understand the round hall, rather than
the square one, as the ancestor shrine. Because of the excavations at
Tillya-tepe, where a cemetery of a
Saka dynasty was found, we may expect to find hidden burials of Parthian kings
in or near Nisa. They may lie, perhaps, in the central area with the temple
massive square may have been a fire temple, or perhaps one of the first examples
of the high towers known as pagodas in China. The height of the recovered stub
does not exclude a taller superstructure, such as known from Buddhist India in
the famous Asoka column. The large square structure in the northern section
looks like a treasury, which could have been a secondary function of the unit.
There is no evidence of it serving as a cemetery.
the fortress of Durnali in the Merv oasis should be mentioned (fig. 10/. It is
located 25 km north of Giaur-kala amid the ruins of a Parthian city, where later
a Seljuk town flourished. The fortress is nearly rectangular, the walls
measuring 183, 173, 160, and 150 meters. The socle was built with pachsa
(raw clay) and the walls, which still stand 7 to 8 meters high, are of brick.
Square towers reinforce the walls, at 10 to 17 meter intervals. The fortress is a typical
example of Parthian fortification.
G. A. Koselenko and V. I. Pilipko, "Issledovannie parfjanskogo
svjatilisca vo okrestnostjach Nisy," Karakumski Drevnosti 2 (Ashkhabad,
1958, 30-35. G. A. Koselenko and G. Gubaev, "Issledovannie parfjanskogo
svjatilisca Mansur-Depe v rannesrednovekogo zamka Ak-Depe," Karakumskie
Drevnosti 14 (Ashkhabad, 1970/, 89--104.
W. Andrae, Hatra, 2 vols. Leipzig,
R. Ghirshman, Terrasses sacrees de Bard-a Nechandeh et Masjid-i Soleiman,
MDAI 45 Paris 1976/; D. Stronach, Pasargadai (Oxford, 1976).
Schlumberger, Der hellenisierte Orient: Die griechische and nachgriechische
Kunst au(3erhalb des Mittelmeerraumes (Baden-Baden, 1969).
M. Gagosidze, "Iz istorii gruzino-iraniskich vzwaimootnosenie,"
Kavkaz i srednaja Azija v drevnosti i srednevekov'e (Moscow, 1981 ), 102-15.
 A. M. Apakidze, G. 1. Gobedcisveli, A. N. Kalandadze, G. A. Lomtatidze, Mzcheta, vol. 1 (Tbilisi, 1958).
P. Tolstov, Po drevnem Del'tam Oksa i faksarta
(Moscow, 1962), pls. 43, 51.
 G. A. Pugacenkova, "Puti razvitija architektury juznogo Turkmenistana pory rabovladenija i feodolizma," Trudy juzno-Turkmenistanskoj archaeologi6eskoj kompleksnoj ekspedizii 4 (1958/, 69-117.
Po drevnem, pls. 58-63.
1. Sarianidi, Baktrisches Gold aus den Ausgrabungen der Nekropolen von
Tillja-Tepe in Nordafghanistan (Leningrad, 1985).
 Pugacenkova, Trudy, 46-50.
Copyright © 1998-2015 The Circle of Ancient Iranian Studies (CAIS)