Author's note: The need for a critical reading of the Süleymaniye defters was first raised in correspondence with Dr. Colin Imber of the University of Manchester. I have profited subsequently from his remarks and, in particular, from the comments of Professor V. L. Ménage of the School of Oriental and African Studies of the University of London. To each I express my gratitude. I am also indebted to Mr. Godfrey Goodwin, to Mr. Geoffrey House of the British Museum and to Dr. Margaret Lyttelton. The defters offer almost unlimited choice of material for study. The present review of the architectural materials is to be supplemented by a review of the decoration and furnishing of Süleymaniye (forthcoming in IJMES); but many problems remain untouched. Transliteration is inevitably somewhat inconsistent. As far as possible I have tried to make the result conform, diacriticals apart, to modern Turkish: where it diverges it is intended to make clearer the original orthography. However, in citing Arabic outside an Ottoman context I have judged it more sensible to adhere to the system of the Encyclopaedia of Islam (2d edition), with the usual modifications.

¹ In particular, his Onaltinct asirda Istanbul hayati (1553–1591), 2d ed. (Istanbul, 1935),Google ScholarHicrî onbirinci asirda Istanbul hayati (1000–1100) (Istanbul, 1930–1931)Google Scholar and Hicrî on ikinci asirda Istanbul hayati (1100–1200) (Istanbul, 1930).Google Scholar

² For a parallel compare the Ashrafiyya madrasa in Jerusalem (van Berchem, M.Corpus Inscriplionum Arabicarum, Vol. II, Syrie du Sud: Jérusalem “Vile” [Cairo, 1923], pp. 358–374),Google Scholar the foundation inscription of which, in the name of al-Malik al-Ashraf Qāyt Bāy, is dated Racab 887'August-September 1482. It was visited by the pilgrim Felix Fabri (Evagatorium II, 24) on 3 August 1483. He found the building a hive of activity, with marble-panellers, inlayers, locksmiths, glaziers and painters still at work. At Süleymaniye, for example, the tiles may not yet have been put up. It has been maintained, with some reservations, by Denny, W. (The Ceramics of the Mosque of Rüsiem Pasha and the Environment of Change [New York and London, 1977], pp. 158–168, 232) that the tile panels for the ḳibla wall of the mosque and the tombs of both Süleyman and Roxelane'Ḫürrem Sulṭān all postdate the foundation inscription and may even postdate the death of Rüstem Paşa in 1561. Ḫürrem SulṬān herself died in 1558, and it is probable that her mausoleum was ready some time previously. But even if the panels were put up after her death tiles in sufficient quantity appear in the defters (Topkapi Saray Archives 45/A, Item 35) for the period ending şa⊂ban 966'June 1559. These must be tiles which were ordered and manufactured. It is not conceivable that tiles for which payment is noted in these accounts could have been an order which was only executed in 1561 or later.Google Scholar

³ Barkan, Ö. L., “L'organisation du travail dans le chantier d'une grande mosquée a Istanbul au XVIe siècle,” Annales, 17 (1962), 1093–1096.CrossRefGoogle Scholar

⁴ Süleymaniye Camii ve Inşaati, I, 16–22.Google Scholar

⁵ Evidently to strengthen the fabric of the dome, as was the case, for example with both the Pantheon and Aya Sofya. The order is thus interesting evidence for the continuity of building practice. CfKaufman, S. A., “O vzaimosvyazyakh rannevizantiiskikh svodchatykh pokrytii s pozdnerimskimi,” Vizantiisky Vremennik, 20 (1961), 184–224.Google Scholar

⁶ Items 1–14 of this list are all stone. Item I, however, was the largest, and not all the others were for building stone.

⁷ CfHinz, W., “Das Rechnungswesen orientalischer Reichfinanzämter im Mittelalter,” Der Islam, 29 (1949), 1–29, 113–141. Some items, for example 33, brass wire/tel-i pirinç, for the domes and windows'ḳubbehā ve cāmhā of the mosque can scarcely have been used.CrossRefGoogle Scholar

⁸ Asgarî, Nuşin, “Roman and Early Byzantine marble quarries of Proconnesus,” The Proceedings of the Xth International Congress of Classical Archaeology, Ankara-Izmir, 23–30'IX'1973 (Ankara, 1978), pp. 467–480. I am grateful for this reference and for observations on the spolia in Haghia Sophia to my colleague, Mr. Geoffrey House.Google Scholar

⁹ Süleymaniye Camii ve Inşaati I, 356,Google Scholar cited after Refik, AhmetHicri onbirinci asirda Istanbul hayati, p. 25. Barkan notes that the omission from the defters of any reference to the Marmara quarries is not crucial, since the quarries for küfek taşi, a kind of coarse sandstone, specifically mentioned in the accounts, at Bakirköy, Yeşilköy and Hazinedār, all to the immediate West of Istanbul, are not mentioned either.Google Scholar

¹⁰ Barkan, “L'organisation du travail,” p. 252,Google Scholar after Refik, AhmetOnalunci asirda Istanbul hayati, p. 21.Google Scholar

¹¹ Tauer, F. “Notice sur les versions persanes de la légende de l'édification d'ayasofya,” Fuad Köprülü Armağani (Istanbul, 1953), pp. 487–494.Google Scholar The stemma of the versions is given by Wittek, P., “Zu den persischen Tārih-i Aya Sofya,” Türkiyat Mecmuast, 14 (1964) 266–270. I am indebted to Professor V. L. Ménage for this further reference. The source was a Byzantine work, the Diegesis, which circa 1000 A.D. was incorporated into a collective work known as the Patria Konstantinoupoleōs.Google Scholar

¹² CfSophiae, Descritio S. (vv. 594 ff);Google ScholarFriedländer, P., ed., Johannes von Gaza und Paulus Silentiarius (Leipzig and Berlin, 1912);Google Scholar and references in Mango, C. and Parker, J., “A Twelfth-Century Description of St. Sophia,” Dumbarton Oaks Papers, 14 (Washington D.C., 1960), 233–245.Google Scholar The extent of the pillage is now somewhat disputed; at least, it no longer seems possible to see in the members of the construction and its decoration the systematic desecration of the great monuments of pagan Antiquity. But the collection of the materials is remarkable enough as a phenomenon in itself. To the same point, there is little evidence that Süleyman was concerned to despoil Christian monuments. An apparent exception was the rather extraordinary incorporation of a conciliar edict of 1116 inscribed on marble slabs and originally erected in Haghia Sophia (Mango, C., “The Conciliar Edict of 1116,” Dumbarton Oaks Papers, 17 [Washington D.C., 1963], 315) into the ceiling of the porch of the mausoleum of Süleyman. According to Pigafetta, however, this occurred in 1567, hence after Süleyman's death.Google Scholar

¹³ Narrative of Travels in Europe. Asia and Africa in the Seventeenth Century by Evliyá Efendi, von Hammer, J., trans. (London, 1834), 1, 68–69. The anecdote, which may well by apocryphal, illustrates the difficulties of the ḥāṣṣa mi⊂marlari in convincing their Imperial patrons that there was more to durable building than their Royal will. The story of the chopped-off hands, Professor Ménage informs me, first appears in Giese's Anonymous Chronicle.Google Scholar

¹⁴ The theft of building materials, incidentally, unlike the illegal confiscation of building land, did not generally invalidate a wakf, much as it was condemned by the ⊂ulamā⊃. It is interesting, moreover, that the acquisition of marble for the mosque of al-Mu⊃ayyad in Cairo (Ibn Taghrīhirdī, Nujūm, Popper, W., trans., History of Egypt, 1382–1469 A.D., Part III, 1414–1422 A.D., [Berkeley and Los Angeles, 1957], p. 41; text VI, p. 360) for which many private houses, and even haunts of vice, were demolished, provoked no protest on the historian's part. What did was the purchase, probably for a derisory sum, of the great doors from the mosque of Sultān Ḥasan and a chandelier from the same foundation. This he qualifies as, inter alia, qillat al-dhawq.Google Scholar

¹⁵ Cf. references in Rogers, J. M., “The Stones of Barqūq,” Apollo Magazine, 103 (01–06 1976), 307–313.Google Scholar

¹⁶ Meinecke, M., “Mamlukische Marmordekorationen in der osmanischen Türkei, Mitteilungen des Deutschen Archäologischen Instituts. Abteilung Kairo, 17, 2 (1971), 209–220.Google Scholar

¹⁷ Badā⊃i⊂ al-Zuhūr, Mostafa, M., ed. (1961), V, 179, 162;Google ScholarWiet, G. trans., Journal d'un bourgeois du Caire. Chronique d'Ibn Iyas, Vol. 2 (Paris, 1960), pp. 156, 173 ff., 176.Google Scholar

¹⁸ Le voyage d'outremer … de Jehan Thenaud suivi de la relation de l'ambassade de Domenico Trevisan auprès du Soudan d'Egypte 1512, Schéfer, Ch., trans. (Paris, 1874), p. 108. The audience hall in the Citadel of Cairo, Pagani asserts, was more beautifully paved than that of the Doge's palace. This is difficult to take seriously. To judge from the extant marble-work in al-Ghawrī's religious foundations the decoration was very stereotyped. It is, of course, conceivable that Pagani is referring to some un-restored decoration from the period of al-Nāṣir Muhammad, the greatest builder at the Citadel of Cairo. But if this were so there is no longer any trace of it.Google Scholar

¹⁹ Mayer, L. A., Islamic Architects and Their Works (Geneva, 1956), p. 66.Google Scholar

²⁰ “Mamlukische Marmordekorationen in der osmanischen Türkei,” p. 210.Google Scholar

²¹ A further indication of the shortage of marble even for panelling is the production of tiles at lznik with designs imitating marble or porphyry, with designs recalling both terazzo and split-marble panelling (Meinecke-Berg, V., “Marmorfliesen. Zum Verhältnis von Fliesendekorationen und Architektur in der osmanischen Baukunst,” Kunst des Orients, 8, 1–2 (1972), 35–59). Such tiles, she observes, play both “architectural” and decorative roles. Thus, the base friezes of the mausoleum of Süleyman the Magnificent and of the miḥrāb of the mosque of Sokollu Meḥmed Paṣa (1571–1572) both employ tiles, the one terrazzo-type (very probably deliberately imitating porphyry) and the other “split-marble,” whereas in earlier periods stone would have been used.Google Scholar

²² Süleymaniye Cammi ve Inşaati, 1, 336–350; II, 11–31. The references and dates of the documents are those given in volume 11, which occasionally show discrepancies with the references in volume 1. Since the former are the texts they must be given priority.Google Scholar

²³ Their location is not specified and their provenance may well have been Cairo or further South. Memphis was a useful source of granite columns, and at least two sets of four are recorded in the Mamlūk monuments of Cairo—in the mausoleum of Qalā⊃ūn (683–684'1284–1285), and in the Qaşr al-Ablaq built by al-Nāşir Muḥammad on the Citadel (713–714'1313–1315). CfCreswell, K. A. C., The Muslim Architecture of Egypt, Vol. II (Oxford, 1959), pp. 191–195, 260–264.Google Scholar

²⁴ Two types of bellūṭ logs are ordered, 400 bellūṭ: çubugu and 60 bellūṭ-i verdinār, the dimensions specified being much the same. The first translates simply as “oak logs”; the second is problematic. For consideration of the timber used in Süleymaniye see Part II of this study, “The Furnishings and decoration of the Mosque” (forthcoming in IJMES).

²⁵ Süleymaniye Camii ye Inşaati, I, 339.Google Scholar

²⁶ Cited by Barkan, , “L'organisation du travail,” p. 48 (text), folio 12b.Google Scholar

²⁷ Iki ḳiṭ⊂a taṣ direk. Nothing is said specifically of granite, and subsequent orders refer invariably to mermer'marble. But there were colossal granite columns at Baalbek. If, for the sake of argument, one of the granite columns now in Süleymaniye is assumed to be from Baalbek this futher shows up the imprecision of the sources.

²⁸ Süleymaniye Camii ye Inşaati, I, 342–343.Google Scholar

²⁹ Ibid., pp. 344–346; II, pp. 23–24, Nos. 44–45. Though Barkan throughout volume I carefully (and in my view rightly) refers to the Kiztaşi Maḥalle the documents refer to the Kiztaşi tout court. But since the Column of Virginity was indisputably porphyry the column brought from that quarter cannot have been it.

³⁰ üserā⊃-i Efrençen nice bin Süleymānī dev. The defters refer to them as irgādān ⊂an ğilmān (or gebrān)-i sefīnehā-i rü⊃esā-i mezkūrīn, without indication of their country of origin. Gebrān is probably not significant here. More important is that they are classified not just as prisoners but as men sentenced to the kürek, and Barkan has judiciously observed that their employment shows a seasonal fluctuation, with a marked fall-off in the late summer when the fleet sailed and the galley-slaves were required for duty. They could thus have been Muslim, as well as Christian, though the term ğilmān is not to be construed without qualifications as “slaves.” If Muslim they would remain free, even if in the galleys; only as Christians would they be automatically enslaved. By the sixteenth century (Heyd, U., Studies in Old Ottoman Criminal Law, Ménage, V. L., ed. [Oxford, 1973], pp. 304–307) penal servitude on the galleys/kürek had come to be applied indiscriminately as a punishment, partly because of the growing need for oarsmen on warships.Google Scholar

³¹ There is no longer sufficient evidence in situ to establish the exact provenance of the Baalbek column(s). Dr. Margaret Lyttelton kindly informs me that the columns of the main porticoes and those fronting the exedras of the temple of Jupiter Helios were of red Aswān granite or of grey granite from the Bosphorus and that the propylon had twelve granite columns. For restorations of these constructions see Th., Wiegand, Baalbek, Vol. I (Berlin and Leipzig, 1921), Plate 33 (exedra) and Plate 41 (propylaea). Granite also seems to have been used for some of the colonnettes of the niches of the articulated facades. These would have been far too small.Google Scholar

³² Süleymaniye Camii ve Inşaati, I, 341 and n. 21.Google Scholar

³³ See Hammer, Narrative of Travel, p. 75. “Besides the square piers which support it, there are, on the right and left sides, four porphyry columns, each of which is worth ten times the amount of the harāj of Miṣr… brought from the capital of Miṣr along the Nile to Alexandria and there embarked on rafts… Four such columns of red porphyry, each fifty cubits high, are to be found nowhere else in the world…” The figure of 50 dirā⊂ is obviously pure invention. The account continues with further misleading details.Google Scholar

³⁴ First ed., (Constantinople, 1924), p. 361.Google Scholar

³⁵ Mango, C., “Antique Statuary and the Byzantine Beholder,” Dumbarton Oaks Papers, 17 (Washington, D.C., 1963), 61.Google Scholar

³⁶ (London, 1971), pp. 227–231, 481.Google Scholar

³⁷ Order No. 16, dated 1 Ṣafar 958'8 February 1551, a list of marbles assembled in Egypt/Cairo by the mi⊂mār Bālī and despatched to Istanbul for the mosque of Süleymaniye, is exceptionally specific. The totals are 121 pieces of black or white marble and 120 pieces of marble of various colours, each total 598 dirā⊂ in length, though the conflation of rectangular panels and roundels makes this figure difficult to interpret. Additionally (unless there is a mistake in the addition and the same marbles are meant) details are given of 203 further pieces: various/mütenevvi⊂ white; somāḵī/porphyry-coloured but here, as in the account of Ibn lyās (see n. 19 above) almost certainly porphyry; ğazalī/gazellecoloured, perhaps (?) tawny; zurzūrī'grey (see p. 73 above); fiṣṭiḵī'pistachio-coloured, hence some shade of green, perhaps cipollino; mersīnī'myrtle-coloured (?); yāsmīnī'jasmine-coloured, evidently some shade of cream; black; zerd-çillī/speckled yellow, probably giallo antico; muḥazza⊂—no congruous sense; sabz'bright green, presumably the green breccia associated with the quarries of the Wādī al-Ḥammāmāt in Upper Egypt, columns of which are still in situ in some of the Mamlūk monuments of Cairo (cf. the mausoleum of the last Ayyūbid ruler of Egypt, al-Ṣāliḥ Najm al-Dīn Ayyūb, built on to the Ṣāliḥiyya Madrasa and the madrasa of al-Nāṣir Muḥammad [Creswell, , Muslim Architecture of Egypt, II, 102, 237]); ḵūlī/(?) read fūlī'bean-coloured, or perhaps a nisba from Ottoman folya/jonquil, hence, perhaps giallo antico; and nawwār al-fu⊃ād. This last is an Arabic-Ottoman calque of Persian dilkushā/heart-rejoicing, which has no obvious sense when applied to marble. For the fiṣṭiḵī, the mersīnī, the yāsmīnī, the black, the zerd-çillī and the sabz there were roundels available but only small slabs. This means that only column fragments had come to light.Google Scholar

³⁸ Cf., Süleymaniye Camii ve Inşaati, II, 115, No. 255, ordering walnut and elm/ḵarāağāç planks, together with beading/parmaḵliḵ; and (p. 114) No. 254, which includes an order for hornbean/gürgen and plane/çinār planks with spade-handles and wood for casks or barrels. Neither order is dated.Google Scholar

³⁹ Ibid., II, 120, No. 261.

⁴⁰ Ibid., 1, 348, n. 22.

⁴¹ Maqrīzī, , Al-Sulūk li-Ma⊂rifat Duwal al-Mulūk, Ziyāda ed., II, 36.Google Scholar

⁴² Maqrīzī, , Al-Mawā⊂iẓ wa'l-l⊂tibār fi Dhikr al-Khiṣaṣ wa'l-āthār (Būlāq, 1853), II, 417.Google Scholar

⁴³ Creswell, , Muslim Architecture of Egypt, II, 248 ff.Google Scholar

⁴⁴ See the references in Layla, ⊂Alī lbrāhīm, “The Great Ḥānqāh of the Amīr Qawṣūn,” Mitteilungen des Deutschen Archäologischen Instituts. Abteilung Kairo, 30, 1 (1974), 37–64.Google Scholar

⁴⁵ Mamboury, E. (Constantinople: A Tourist's Guide [Constantinople, 1924], p. 361)Google Scholar asserts, doubtless following Gyllius, who arrived in Istanbul in January 1550, that the columns of the atrium were all brought from the Hippodrome (cited Ebersolt, J., Constantinople byzantine et les voyageurs du Levant [Paris, 1918], pp. 78–80). He reports that the Hippodrome still had a portico standing with seventeen white marble columns, all with bases and capitals, which were removed “to make a hospice for the Sultan.” This is not incompatible with Mamboury's conclusion. The general meaning of Ottoman ⊂imāret is “hospice,” but the vast enterprise which Süleyman had undertaken is regularly referred to in the documents as the ⊂lmāret-i ⊂āmire, and, by a comprehensible metonymy, it is also used for the mosque. If the columns were used, therefore, it could as well have been inside the mosque as in the atrium. But they were obviously not all used, and if any of them were used in the atrium they must have been cut down.Google Scholar