NAQŠ-e ROSTAM, a perpendicular cliff wall on the southern nose of the Ḥosayn Kuh in Fārs, about 6 km northwest of Persepolis. The site is unusually rich in Achaemenid and Sasanian monuments, built or hewn out from the rock. The Persian name “Pictures of Rostam” refers to the Sasanian reliefs on the cliff, believed to represent the deeds of Rostam.

Achaemenid Period. The most important architectural remains are the tower called Kaʿba-ye Zardošt (Kaʿba of Zoroaster,Ar. kaʿba “cube, sanctuary”) and four royal tombs with rock cut façades and sepulchral chambers.

(1) The Kaʿba-ye Zardošt is a massive, built square tower, resting on three steps (7.30 x 7.30 x14.12 m) and covered by a flat pyramidal roof (Stronach, 1967, pp. 282-84; 1978, pp. 130-36; Camb. Hist. Iran II, pp. 838-48; Schmidt, pp. 34-49). The only opening is a door. But on all four sides there is a system of blind windows in dark grey limestone, set off by the yellow color of the general structure, between the reinforced corners, and the walls are covered with staggered rectangular depressions. Both systems have no other purpose than to relieve the monotony of the structure. A frieze of dentils forms the upper cornice (FIGURE 1). A staircase of 30 steps, eight of which are preserved, led to the door (0.87 x 1.75 m) in the upper part of the north wall. Originally, the two leaves of a door opened into an almost square room (3.72 x 3.74 x 5.54 m) without any architectural decoration and no provisions for lighting (Schmidt, p. 37).

There is an analogous, though much more decayed, structure, called Zendān-e Soleymān (lit. prison of Solomon), in Pasargadae (Stronach, 1978, pp. 117-37; 1983, pp. 848-52). Its stone technique does not yet show traces of the toothed chisel (Stronach, 1978, p. 132), and the building can thus be dated to the last years of Cyrus II the Great (r. ca. 558-530 BCE), whereas due to chisel marks the Kaʿba-ye Zardošt can be dated to the early years of Darius I (r. 522-486), around 500 BCE. The Achaemenid structures do not have exact prototypes, but their plan is comparable with those of the earlier Urartian tower temples (Stronach, 1967, pp. 278-88; 1978, pp. 132-34).

On the Kaʿba-ye Zardošt, three exterior sides bear the famous inscription of Shapur I. (r. 241-72 C.E.). The Res gestae divi Saporis (ŠKZ) was added in Greek on the south wall, in Sasanian Pahlawi (Parsik) on the east, and in Parthian (Pahlawik) on the west (Back, pp. 284-371), while the north wall with the entrance has remained empty. Beneath the Parsik version on the east wall, the high priest Kirdīr had his own inscription incised (Sprengling, pp. 37-54; Chaumont, pp. 339-80; Gignoux, pp. 45-48). Evidently, in Sasanian times the Kaʿba-ye Zardošt—like the tower at Paikuli with the inscription of Narseh (r. 293–302; cf. Humbach and Skjaervø)—served, in addition to other functions, as memorial. Perhaps the two towers in Naqš-e Rostam and Pasargadae already had a similar significance in Achaemenid times, albeit this cannot have been their main function.

In Kirdīr’s inscription the Kaʿba-ye Zardošt is called “bun-xānak.” W. B. Henning proposed the translation “foundation house,” and concluded that the tower was of central religious significance. He suggested that the empty high room was destined “for the safe keeping of the records of the church and even more for the principal copy of the Avesta” (Henning). Though other translations of “bun-xānak” have been discussed (Gignoux, pp. 28-29 n. 61), it seems the most convincing interpretation that these two towers served as depositories. The lack of any provision for the ventilation of a fire excludes the towers’ use as fire temples (Stronach, 1978, pp. 134-35). Their staircases were designed “for the solemn ascent and descent of persons who in some manner attended the sacred structure” (Schmidt, p. 41). They indicate that the towers did not serve as royal tombs (Stronach, Camb. Hist. Iran II, p. 849 n. 2), because those have entrance walls that are smoothed beyond their facades, down to the original ground, to make them inaccessible.

R. N. Frye (1974, p. 386) first expressed the opinion that “the intention was . . . to build a safety box for the paraphernalia of rule in the vicinity of Persepolis as had been done at Pasargadae,” though E. F. Schmidt (p. 44) had dismissed the interpretation of the Kaʿba-ye Zardošt as depository. But Plutarch (46-after 119 C.E.) mentions in Artoxerxes 3 that at Pasargadae one temple belonged “to a warlike goddess, whom one might conjecture to be Athena” (Sancisi-Weerdenburg, p. 148). At this sanctuary the Achaemenid kings were crowned. During the coronation ceremony the new monarch took a very frugal meal, and was dressed in the robes which Cyrus the Elder wore before assuming kingship. H. Sancisi-Weerdenburg was the first to identify the Zendān-e Soleymān as Plutarch’s temple (Gk. hieron). Consequently, she interpreted this building, as well as the Kaʿba-ye Zardošt, as “coronation tower.” Her view that these towers had dynastic functions, rather than a purely religious significance and definitely no funeral purposes, has become widely accepted, though her suggestion that a sacred fire was also kindled in these towers can no longer be upheld.

(2) The Royal Tombs. In the cliff wall four monumental tombs are cut out from the native rock (Schmidt, pp. 80-107). The oldest tomb (Tomb I) has inscriptions that assign it to Darius I. The other three (Tomb II-IV) can only tentatively be attributed to Xerxes (eastnortheast of Darius I), Artaxerxes I (westsouthwest of the tomb of Darius I) and Darius II (westernmost).

The four monuments follow the same pattern. But it is completely different from that of the older tomb of Cyrus the Great at Pasargadae, which is a built structure consisting of a stepped platform and a tomb with a gabled roof. The model was first used for Darius I (FIGURE 2) and has no exact prototypes in the Near East, Egypt or Greece, though the stone technique is Urartian in origin (Calmeyer, 1975, pp. 101-7; Gropp, pp. 115-21; Huff, 1990, pp. 90-91).

The rock tomb is characterized by the contrast between a cruciform composition in relief on the exterior wall and a very simple interior of chambers and grave cists. The center of the relief ensemble is a facade that represents the front of a palace with four engaged columns. On this architectural component rests a throne bench (Gk. klinē, OPers. gathu in inscription DNa) that is supported by 30 representatives of the empire’s peoples. The throne bench in turn serves as the platform of a religious scene with king, fire altar, and divine symbols (FIGURE 3).

The architectural register recalls the palace of the living monarch because the portico’s dimensions on the tomb of Darius I. are almost identical to those of his palace on the terrace of Persepolis (Schmidt, p. 81). A significant feature is the use of engaged columns, which appear on his tomb for the first time in rock architecture. The so-called Median rock tombs, which are imitations of the Achaemenid monuments, do always show free standing columns (von Gall, 1966, pp. 19-43; 1973, pp. 139-154; 1988, pp. 557-82; “Dokkān”); the exception is the tomb of Qizqapan, where half columns have been placed on the rear of the antechamber (von Gall, 1988, pl. 23). But at many tombs in the Median province, the originally freestanding columns have collapsed under the pressure of the superimposed rock. Consequently, there was not only the esthetic reason of creating the illusion that the antechamber’s front side and back wall were on the same level. More important were statical considerations. The architects and sculptors of the royal tombs used engaged columns because they could withstand the rock pressure despite their high slender shape.

In the middle register, the mighty throne bench with its 30 armed carriers does not show a realistic scene, and is not considered pictorial evidence for the supposition of real processions on the roofs of Achaemenid palaces (Schmidt, p. 80). It rather is a simile of the Achaemenid empire, the throne bench of which is supported by its peoples, dressed in their distinctive costumes and headgears (Schmidt, pp. 108-118). On the tombs of Darius I in Naqš-e Rostam and that of Artaxerxes II (r. 404-359 BCE) in Persepolis, inscriptions describe the peoples’ order, and this order seems to correspond with the official geographical records of the empire’s extension (Calmeyer, 1982, pp. 109-123). According to P. Goukowsky (p. 223; cf. Calmeyer, 1982, p. 113 fig. 3) the empire was divided in three concentric zones: Persians, Medians and Elamites live in the inner circle. An axis is leading from the center to the east, listing Parthians, Arians, Bactrians, Sogdians, and Chorasmians. Then the enumeration turns southeast, naming Drangians, Arachosians, Sattagydians (Thataguš), Gandharans, and Indians and reaches Central Asia, where the haoma-venerating Scythians (FIGURE 6) and pointed-hat Scythians already inhabit the periphery. On a second axis leading to the south the Babylonians, Syrians, Arabians, and Egyptians (Mudraya) are aligned, whereas on a third axis to the northwest the Armenians, Cappadocians, Lydians (Sparda), and Ionians are represented. Finally in the western periphery there live the Scythians beyond the Sea, the Thracians (Skudra), and the Petasos bearing Ionians. The Libyans (Putaya) and the Ethiopians (Kušiya) roam the empire’s southernmost countries. Two men stand outside the throne bench, and their hands help lifting the platform which is slightly elevated above the ground. They are a Makan (Maka, i.e., Oman and probably also the region on the Persian side of the Gulf) and a Carian (Karka). P. Calmeyer (p. 120) has convincingly argued that their exceptional corner positions reflects that these two peoples inhabit the south and the west corners of the empire, at the shore of the ōkeanos which in antiquity was believed to flow around the inhabited earth (Gk. oikoumenē). All men (Schmidt, figs. 39-50), with the exception of the Babylonian (ibid., fig. 50 no. 16), are wearing weapons, mostly daggers and swords, and some also pairs of javelins. Bearing arms in the presence of the monarch was a sign of honor and trust, so that the unarmed Babylonian represents an act of deliberate humiliation. Since Xerxes (r. 486-465 BCE) probably supervised the final work on the tomb of his father Darius I (Schmidt, pp. 116-18 part. 117), this humiliation is likely to reflect to repeated rebellions of the Babylonians against him as well as against his father.

The scene in the top register has religious significance. The king is standing on a three-stepped platform, his left resting on a bow, while his slightly lifted right hand points to the winged symbol hovering above the scene (FIGURE 4, FIGURE 5). Since the late 19th, early 20th century, the winged ensign with a human figure, emerging from a circle, has been understood as a representation of Ahura Mazdā (Root, pp. 169-79), and recent attempts to interpret this symbol as the royal genius Frawahr have been rejected. The king faces a blazing fire altar, though he stands at a considerable distance, whilst the ensign of a disc with inscribed crescent is hovering in the upper right corner. In general, scholars agree that this scene shows how the king is worshipping the holy fire. But the gesture of the king’s right hand corresponds in all details with that of the right hand of the Ahura Mazdā symbol. The representation thus stresses the close connection between the king and Ahura Mazdā, whose will is decisive for the king’s actions. This interpretation is supported by the Achaemenid royal inscriptions, which are directly related to the reliefs.

On tomb I, Darius I wears a headdress (Gk. kidaris) with an upper rim of sculptured stepped crenellations (FIGURE 4). Reliefs on the jambs of the southern doorway in Darius’s Palace (Tilia, pp. 58-59) indicate that this was the personal crown of Darius, which was also worn by Xerxes as long as he was crown prince (von Gall, 1974, pp. 147-51). On Tomb II, which is ascribed to Xerxes, in the king’s crown (FIGURE 5) the rest of a sculptured crenellation is visible (von Gall, 1974, pl. 134 no. 2; 1975 fig. 3), suggesting that this monument was completed before he became the absolute monarch (von Gall, 1974, p. 151). The representations of this late time show a straight cylindrical crown without any decoration. All succeeding rulers of the Achaemenid dynasty adopted this shape, allowing only minor deviations (von Gall, 1974, pp. 150-60; 1975, pp. 222-24).

Another invariable detail of the royal tombs is the discoid symbol hovering in the upper right corner. The inscribed crescent indicates its Assyrian origins. While it represents the moon god Sin in Assyrian art, on the Achaemenid tombs its meaning is difficult to comprehend. Opinions differ whether the symbol has to be interpreted as lunar or solar (cf. Root, pp. 177-78), and there are no written sources to corroborate either view. E. F. Schmidt (p. 85) interpreted the sign as a symbol of Mithra. But the Persian moon god Māh is relatively well documented in the imagery of the Achaemenid seals. In the central panel above the fire altar scene of the rock tomb of Qizqapan, this type of moon god is also represented (von Gall, 1988, pp. 571-72). These images, in connection with other, though scanty, pictorial evidence (von Gall, 1988, p. 572 n. 55), suggest that the moon played a certain role in Achaemenid concepts of death and afterlife.

On the tomb of Darius, the framework of the throne bench shows three superimposed figures on each side. On the left, two dignitaries are inscribed as the lance bearer Gobryas (Gaubaruwa) and the bearer of the royal battle-ax Aspathines (Aspačina), while the lowest man is an unnamed guard (Schmidt, pp. 86-87). On the right, three unarmed men are clad in the long Persian garment. Their gesture of raising a part their upper garment to the mouth has been interpreted as an expression of mourning, comparable to the Greek custom (Schmidt, p. 87). More recently, scholars have suggested that this gesture captures the imperative of ’do not pollute the holy fire’ (Hinz, p. 63 n. 4; cf. Seidl, p. 168) or shows respect for the king’s majesty (Root, p. 179), but both alternatives seem less convincing. Additional figures are on the side walls of the recesses into which the tomb facade was carved. On the left, there are three superimposed panels with guards holding long lances. On the right, three mourners who need be considered either courtiers or members of the royal family (Schmidt, p. 87) stand above each other.

Two larger cuneiform inscriptions, as well as legends with the names of Darius I, of his two supreme commanders, and of the 30 bearers of the throne bench, are found in the facade of Tomb I. One is in the top register, to the left of the king (DNa, cf. FIGURE 4), and the other (DNb) stands in the architectural register, on three of the five panels between the half columns of the portico. Both are written in three languages, but DNa in Old Persian, Elamite, and Babylonian (Weissbach, pp. 86-91), and DNb in Old Persian, Elamite and Akkadian (Hinz, pp. 52-62 including R. Borger’s edition of the Akkadian version). In the Seleucid period, an Aramaic version was added to DNb below the Elamite text (Frye, 1982).

In stark contrast to the rich architectural decoration of the façade, the interior is bare of any architectural and figural elements. The general layout is also best demonstrated with the tomb of Darius I: A long vestibule is running parallel to the facade, and three doors in the back wall of this vestibule are leading to three separate barrel-vaulted tomb chambers. In each tomb chamber, a trough-like cavity was hewn into the solid rock to hold a probably wooden sarcophagus or klinē. These cists were sealed with monolithic lids after the deposition of the corpses, but nothing has remained of the original interments.

The combination of an oblique corridor and burial chambers with cists was preserved in the other three tombs, assigned to Xerxes (Tomb II), Artaxerxes (Tomb III), and Darius II (Tomb IV). Yet they show inferior craftsmanship, because the chambers are not running axially, but obliquely to the facade. At Persepolis, the interior organization of the two tombs is also identical.

(3) Other architectural remains. In the Center Test of his 1936 and 1939 excavations, E. F. Schmidt found a building (Schmidt, pp. 10 and 64). In the West Test, he discovered remains of two mud-brick buildings, as well as evidence of an enclosure of the royal tombs (ibid., pp. 10, 54-55). In the west of the cliff, a polygonal cistern (diam. 7.20 m) hewn out from the native rock was excavated (ibid., pp. 10, 65).

The Sasanian Period. A fortified enclosure ran around the major part of the sculptured cliff, and its west and east ends were abutting with the rock. Seven semicircular towers strengthened this structure (Schmidt, pp. 55-58, figs. 2, 4; cf. Trümpelmann, p. 41, fig. 68, drawing by G. Wolff). On the slope of the Hosayn Kuh, there are two cut rock structures in the shape of a čahārṭāq. They are generally assumed to be Sasanian fire altars, but D. Huff (1998, p. 80 pl. 10a; “Fārs,” pp. 353-54 pl. 3) identifies them as astōdāns.


M. Back, Die sassanidischen Staatsinschriften, Acta Iranica 18, Leiden, 1978.

P. Calmeyer, “Zur Genese altiranischer Motive III: Felsgräber,” AMI NF 8, 1975, pp. 99-113.

Idem, “Zur Genese altiranischer Motive VIII: Die ’statistische Landcharte des Perserreiches’ I-II,” AMI 15, 1982, pp. 105-187, and 16, 1983, pp. 141–222.

M. L. Chaumont, “L’inscription de Kartir à la ‘Ka’bah de Zoroastre’: Texte, traduction, commentaire,” JA 248, 1960, pp. 339-80.

A. Demandt, “Studien zur Kaaba-i-Zerdoscht,” Archäologischer Anzeiger, 1968, pp. 520-40.

DNa = Inscription a of Darius I from Naqš-e Rostam; for the text, see Weissbach, pp. 86-91.

DNb = Inscription b of Darius I from Naqš-e Rostam; for the text, see Hinz, pp. 522 and 62; and Frye, 1974.

R. N. Frye, “Persepolis again,” JNES 33, 1974, pp. 383-86.

Idem, “The ’Aramaic’ Inscription on the Tomb of Darius,” Iranica Antiqua 17, 1982, pp. 85-90.

H. von Gall, “Zu den ’medischen’ Felsgräbern in Nordwestiran und Iraqi Kurdistan,” Archäologischer Anzeiger, 1966, pp. 19-43.

Idem, “Neue Beobachtungen zu den sog. medischen Felsgräbern,” in Proceedings of the IInd Annual Symposium on Archaeological Research in Iran, 1973, Tehran, 1974, pp. 139-54.

Idem, “Die Kopfbedeckung des persischen Ornats bei den Achämeniden,” AMI N.F. 7, 1994, pp. 145-61.

Idem, “Die großkönigliche Kopfbedeckung bei den Achämeniden,” in Proceedings of theIIIrd Annual Symposium on Archaeological Research in Iran 1974, Tehran, 1975, pp. 219-32.

Idem, “Das Felsgrab von Qizqapan: Ein Denkmal aus dem Umfeld der achämenidischen Königsstraße,” Bagdader Mitteilungen 19, 1988, pp. 557-82.

Idem, “Dokkān-e Dāwūd,” EIr VII, pp. 472-74.

Ph. Gignoux, Les quatres inscriptions du mage Kirdir: Textes et concordances, Studia Iranica cahiers 9, Paris, 1991, pp. 45-48.

P. Goukowsky, Essai sur les origins du mythe d’Alexandre 336-270 av. J.-C.: I – Les origines politiques, Nancy, 1978.

G. Gropp, “Urartäische Miszellen,” AMI 22, 1989, pp. 103-124.

W. B. Henning, The Inscription of Naqš-I Rustam, Corpus Inscr. Iran. 3,3,2, London, 1957, introduction.

W. Hinz, Altiranische Funde und Forschungen, Berlin, 1969.

D. Huff, “Das Grab des Doğubayazıt: Seine Stellung unter den urartäischen und iranischen Felsgräbern,” Türk Tarih Kongresi 10, 1990, I, pp. 87-95; proceedings of a conference held in Ankara in 1986. Idem, “Fire Altars and Astodans.” in The Art and Archaeology of Ancient Persia: New Light on the Parthian and Sasanian Empires, ed. V. Sarkosh Curtis et al., London 1998, pp. 74-83.

Idem, “Fārs V. Monuments,” EIr IX, pp. 351-56.

H. Humbach and P. Skjaervø, The Sassanian Inscription of Paikuli, parts 1-3,2, Wiesbaden, 1978-1983.

F. Krefter, “Achämenidische Palast- und Grabtüren,” AMI N.F.1, 1968, pp. 99-113.

M. Cool Root, The King and Kingship in Achaemenid Art, Acta Iranica19, Leiden 1979.

H. Sancisi-Weerdenburg, “The Zendan and the Kaʿbah,” in Kunst, Kultur und Geschichte der Achämenidenzeit und ihr Fortleben, ed. H. Koch and D. N. Mackenzie, AMI Ergänzungsband 10, 1983, pp. 145-51.

E. F. Schmidt, Persepolis: III – The Royal Tombs and other Monuments, Oriental Institute Publications 70, Chicago 1970.

U. Seidl, “Naqš-i Rustam,” RlA, IX, pp. 165-68.

ŠKZ = Ph. Huyse, Royal Inscriptions with their Parthian and Greek Versions: Texts I – Die dreisprachige Inschrift Šābuhrs I an der Kaʿba-i Zardušt (ŠKZ), Corpus Inscr. Iran. 3,1,1,1.2, London, 1999.

M. Sprengling, Third Century Iran: Sapor and Kartir, Chicago, 1953.

D. Stronach, “Urartian and Achaemenian Tower Temples,” JNES 26, 1967, pp. 278-88.

Idem, Pasargadae: A Report on the Excavations Conducted by the British Institute of Persian Studies from 1961 to 1963, Oxford, 1978.

Idem, “The Median and Achaemenian Periods,” Camb. Hist. Iran II, pp. 838-55.

A. B. Tilia, Studies and Restorations at Persepolis and other Sites of Fārs,vol. II, Istituto Italiano per il Medio ed Estremo Oriente: Reports and Memoirs 18, Rome, 1978.

L. Trümpelmann, Zwischen Persepolisund Firuzabad: Gräber, Paläste und Felsreliefs im alten Persien,ed. M. Abka’i-Khavari and D. Berndt, Mainz, 1991.

F. H. Weissbach, Die Keilinschriften der Achämeniden, Leipzig, 1911.

February 20, 2009

(Hubertus von Gall)

Originally Published: February 20, 2009

Last Updated: February 20, 2009