^1. Takht-i-Sulayman was visited in October 1937 by the eighth Architectural Survey expedition of the American Institute for Iranian Art and Archaeology The ruins were photographed by Arthur Upham Pope, the Director of the Institute, and Stephen Nyman, and surveyed and recorded by Donald Wilber, John McCool, Dr J. Christy Wilson and Clair Armstrong. A first account of the results was given in the London Times of 18 February 1938 and additional photographs were published in the Illustrated London News of 26 February 1938: both articles were by Arthur Pope. The history of the place, called Shiz by the Arabs and Phraaspa in Parthian times, is discussed by Mary Crane: ‘II, The Historical Documents’ in the Bulletin of the American Institute for Iranian Art and Archaeology, December 1937. In this same publication a general description of the architectural remains is given; Donald Wilber, ‘3, Description of the Extant Structures’. In the preparation of the present article the author received valuable suggestions from Professor E. Baldwin Smith of Princeton University.

^2. The buildings are shown in FIG. 2 as A and B.

^3. SirRobert Ker, Porter in Travels in Georgia, Persia, Armenia, Ancient Babylonia, c. (London, 1821–22), 2, Google Scholar p. 560 says ‘to the southwest, a large square building of hewn stone, with a wide columned portal of a hard red kind of marble. Part of the shafts and torus of the columns were still there, with some fragments of a curious fretwork carved on the same sort of stone’. Henry Rawlinson in ‘Notes on a Journey from Tabriz, through Persian Kurdistan, to the ruins of Takhti-Soleiman, and from thence by Zenjan and Tarom, to Gilan, in October and November, 1838, with a memoir of the site of the Atropatenian Ecbatana’, Journal of the Royal Geographical Society, x (1841), p. 52 reports ‘a small square enclosure of four walls rudely built of unhewn stone, near the south-western face of the fortifications; part of the left hand column of the gateway is still standing, formed of huge blocks of a dark-red stone, which are cut into the shape of the outer half of an octagon, and are also carved with an ornamental pattern; two fragments of a shaft are standing erect in front of the gateway; two others are lying on the ground near it; and within the walls there are also two bases or capitals; for it is not easy to distinguish which; all formed of the same dark-red stone, that is not to be met with in any other part of the ruins’. A. Houtum-Schindler in his ‘Reisen im Nordwestlichen Persien 1880-82’, Zeitschrzft der Gesellschaft fur Erdkunde mc Berlin, 1883, XVIII, p. 328 describes the structure: ‘Somewhat further west lay the ruins of a hall directed from north to south. The square stones of the gate were soft red sandstone . . . and on two round columns, which measured 0.8om. in diameter, I saw tracea of arabesques’

^4. This mosaic is published in Briggs, W.A. PompeiQn Decoration London, 1911.Google Scholar It is shown on the third of the unnumbered plates. The author suggests that the mosaic dates between 200 and 80 B.C.

^5. The simple torus used as a column capital has a long history in the Near East, appearing at such different sites as Nineveh, Khorsabad, Tell Halaf and Susa.394

^6. A comparable treatment is found in the early Sasanian fire temple, possibly little more than a century later in date than the building under consideration, at Robat-i-Safid in Khorasan. The plan is given in FIG. 10. Here the central dome chamber is flanked north and south by a long narrow room. These rooms have a series of beam holes in their walls which show that they were originally divided into three storeys. Slanting lines of beam holes indicate the original position of the wooden stairs.

^7. Walter, Andrae, Hatra nach Aufnahmen von mitgliedern der Assur-expedition der Deutschen Orient-Gesellschaft, 1, Leipzig 1908,Google Scholar pls. 11 and x and vol. 11, Leipzig, 1912, 221, PIS. VII, IX, XI

^8. Walter, Andrae und Heinz Lenzen, Die Partherstadt Assur, Leipzig 1933,Google Scholar fig. 20.

^9. Plans of a number of these temples are given in Butler, H.C. ‘Nabataean Temple Plans and the Plans of Syrian Churches’, in Studien zur Kunst des Ostens, Vienna, 1925, pp. 9–16.Google Scholar

^10. lo In most cases the columns were not actually found in situ but lying in the nearby vicinity; see Butler, H.C. Princeton Expedition to Syria, division 11, Ancient Architecture, section A, p. 374.Google Scholar

^11. See Butler, Stdim zur Kumt des Ostm, p. 2.Google Scholar

^12. There is another Parthian structure whose plan is almost as close to the Nabataean form but a drawing of it has not yet been published. The structure is a mausoleum the principal chamber of which is square in plan and with four cruciform piers forming an interior square. It was discovered by Soviet excavators at Nesa near ‘Ashqabad and is referred to in an editorial note in A Survey of Persian Art (forthcoming), I, p. 444.

^13. There is no reason for examining this evidence at any length. Monuments which are either known to have had or are thought by some scholars to have had domes of wood include the church at Kalcat Sim‘an, the cathedral at Bosra, the church of Constantine at Antioch and the Dome of the Rock at Jerusalem.

^l4. Ernst, Herzfeld, Archaeological Htstory of Iran, London, 1935 p. 88–89.Google Scholar

^l5. Ugo Monneret de, Villard, ‘The Fire Temples’, in the Bulbtin of the Amkan An additional subject remains for consideration. Institute for Persian Art and Archaeology, December, 1936 p. 175–184.Google Scholar