^{page 469 note 1} In presenting these studies, it is my pleasant duty to record my thanks to the office of the Turkish Prime Minister for giving me access to the state archives in Istanbul; to the Director and staff of the archives, and especially B. Midhat Sertoğlu of the General Directorate, for their unfailing courtesy and co-operation; to Professor M. Tayyib Okiç of the Faculty of Divinity of the University of Ankara, who first initiated me into the mysteries of Siyāqat and was always ready at hand with assistance and advice during my work in the archives; to my colleague Professor P. Wittek for many helpful criticisms and suggestions throughout the preparation of these studies; to my wife, who did much of the work involved in preparing the statistical sections and also in preparing the manuscript for the press.

^{page 469 note 2} The first part of this paper was read to the 23rd International Congress of Orientalists in Cambridge, in August, 1954.

^{page 470 note 1} ‘ The Ottoman Archives as a Source for the History of the Arab Lands ’, JRAS, 10, 1951, 139–155Google Scholar; Notes and Documents from the Turkish Archives, Jerusalem, 1952.Google Scholar

^{page 470 note 2} Kun, T. Halasi, ‘ Avrupadaki Osmanli Yer Adlari Üzerinde Araştirmalar ’, Türk Dili ve Tarihi hakkinda Araştirmalar, Ankara, 1952, 82 ff.Google Scholar

^{page 470 note 3} That is, the south-western part of Albania. An edition of this important text by Professor Halil Inalcik of Ankara is due to appear shortly. See Inalcik, Halil, ‘ Timariotes Chrétiens en Albanie au XV. siècle d'après un registre de timars ottoman ’, Mitteilungen des Österreichischen Staatsarchivs, 4, 1952, 118 ff.Google Scholar

^{page 470 note 4} A descriptive list of these registers will be found in ‘ The Ottoman Archives …’, JRAS, 154 ff.

^{page 471 note 1} An example is given in appendix I below.

^{page 471 note 2} A number of qānūnnāmes of Palestine and Syria have been published. Turkish texts will be found in Barkan, Ömer Lutfi, XV ve XVI inci Asirlarda Osmanli Imparatorluğunda ziratî ekonominin hukukî ve malî esaslari, Vol. 1, Kanunlar, Istanbul, 1943Google Scholar, and French translations in Mantran, R. and Sauvaget, J., Règlements Fiscaux Ottomans, Les Provinces Syriennes, Paris, 1951.Google Scholar For older German renderings of some qānūnnāmes see Hammer, J. von, Des Osmanischen Reichs Staatsverfassung und Staatsverwaltung, Vienna, 1815, vol. 1.Google Scholar See further ‘ The Ottoman Archives ’, 144–5. The qānūnnāmes for Syria and Palestine make frequent allusions to the qānūnnāmes of Qā'it Bāy, which in many respects they confirm. A statement of Mamluk fiscal practice in Syria will be found in the Qānūn al-Bilād asẖ-Sẖāmīya given by Nuwairī, , Nihāyat al-Arab, 8, 255 ff.Google Scholar

^{page 472 note 1} Thus in effect maintaining the two Niyābas of Safed and Gaza, into which Palestine was divided under the Mamluks.

^{page 472 note 2} On these officers and their functions see Gibb, H. A. R. and Bowen, Harold, Islamic Society and the West, 1, 1, Oxford, 1950, 51 and 145 ff.Google Scholar

^{page 472 note 3} In several registers the nāḥiye around Safed is called Jira.

^{page 474 note 1} Christian monks in Jerusalem, however, appear to be exempted in some registers though not in others. On the exemption of monks in the Ottoman Empire see Lemerle, P. and Wittek, P., ‘ Recherches sur l'histoire et le statut des monastères athonites sous la domination turque ’, Archives d'histoire du droit oriental, 3, 1948, 468–9.Google Scholar

^{page 474 note 2} al-Ḥalabī, Ibrāhim b. Muḥammad, Multaqā al-Abḥur, Cairo, n.d. (? 1910), 126Google Scholar; D'Ohsson, M., Tableau Géneral de l'Empire Ottoman, 5, Paris, 1824, 23.Google Scholar (cf. Margẖinani, ii, 137, al-Wiqāya, 246, Qudūri, Muḵẖtaṣar, 131, quoted by Strauss, E., Tōledōth ha- Yehūdīm be-Miṣrayyim ve-Sūriya, 2, Jerusalem, 1951, 263, n. 3.)Google Scholar

^{page 475 note 1} In the absence of any data on the number of persons to a household, this figure is of course conjectural, ca. 5–7 seems a fair average.

^{page 476 note 1} These figures will be discussed in greater detail in the next article.

^{page 479 note 1} On the Jewish inhabitants see further Notes and Documents, 5 ff.

^{page 479 note 2} On the Döger Turcomans see Sümer, Faruk, ‘ Dögerlere dair ’, Türkiyat Mecmuasi, 10, 1953, 139 ff.Google Scholar

^{page 479 note 3} So I interpret the quarter named in the registers. The name already occurs in Mamluk times—see ad-Din, Mujir, Al-Uns al-Jalīl, Cairo, 1283, 419Google Scholar; French translation in Sauvaire, H., Histoire de Jérusalem et d'Hébron, Paris, 1876, 209.Google Scholar On some Bashkirs in the Mamluk service see Togan, Zeki Velidi, x2018; Baskirt ’, Islam Ansiklopedisi, 2, 329–330.Google Scholar

^{page 480 note 1} cf. Ayalon, D., ‘ The Wafidiya in the Mamluk Kingdom ’, Islamic Culture, 1951, 89 ff.Google ScholarPoliak, A. N., Feudalism in Egypt, Syria, Palestine, and the Lebanon, 1250–1900, London, 1939, 9.Google Scholar

^{page 480 note 2} That is to say, the Sultan granted a brevet (berāt) on production of a certificate (tedẖkere) from the Beylerbey. (See J. Deny, ‘ Timar ’, E.I.) In the earlier timar registers transfers (taḥvīl) of fief-holders to new fiefs were noted, with the date and details of the transfer. An example is given in appendix III. Separate registers of taḥvīls are preserved in the archives in Istanbul, covering the period 1024–1292. The papers of the taḥvīl qalemi go back to 895 A.H.

^{page 481 note 1} Qawānīn-i Āl-i 'Osmān, Constantinople, 1864.Google Scholar (cf. Belin, K. in JA 6th series, 15, 1870, 241–244 ff.Google Scholar and Tischendorf, P. A. von, Das Lehnswesen in den Moslimischen Staaten, Leipzig, 1872, 88–90.)Google Scholar

^{page 481 note 2} The holding recorded in any one place may be only a portion of the fief held by the individual concerned. Where this is so, the heading ‘ zi'āmet ’ or ‘ timar ’ is preceded by the Arabic preposition ‘an, which here has the meaning of‘ a part of ’. In the same way, when the revenues of a village or mezra'a included in the enumeration of a fief are shared with some other beneficiary, the preposition ‘an is prefixed to the entry, and the share (ḥṣṣe) of the fief-holder is indicated. Examples of this may be seen on Plate III.

^{page 481 note 3} On these officers, who were concerned with the registration of fiefs in the Vilāyet, see Gibb and Bowen, 149 ff.

^{page 482 note 1} No total is given in the register. This figure represents the sums of the individual sources of income listed.

^{page 482 note 2} No totals are given for these two zi'āmets. The figures given above were obtained by adding together the amounts of the individual sources of revenue of each. The register gives for the combined revenue of both zi'āmets the somewhat higher figure of 37,108.

^{page 483 note 1} cf. Hammer, , Staatsverfassung, 1, 203–5.Google Scholar

^{page 484 note 1} The word dīmūs does not, to my knowledge, occur in any Islamic document of the pre- Ottoman period, and even in the Ottoman period it is limited to Syria. The word is obviously of Greek origin, and is probably a corruption of Δημόσια, which occurs in both Greek and Jewish Aramaic texts of the pre-Arab period in relation to money taxes in Palestine (examples in Cange, Du, Glossarium ad scriptores mediae et infimae graecitates, Lyons, 1688, 1, 288Google Scholar, Δημόσια; Jastrow, M., A Dictionary of the Targumim, etc., New York, 1926, 1, 300Google Scholar; See further Goldschmidt, L., ‘ Les impôts et droits de douane en Judée sous les Romains ’, RÉJ, 34, 1897, 204–5Google Scholar; Gulak, A. in Magnes Anniversary Book, Jerusalem, 1938, p. 98Google Scholar of the Hebrew text; Avi-Yonah, M., Biyyemê Roma uvizantion, Jerusalem, 1946, 57 and 60).Google Scholar It would thus appear that this Greek term remained in current use in Syria throughout the Islamic period, but was never accepted into the technical language of the jurists. Only after the Ottoman conquests, when for the first time we have administrative as distinct from juristic and literary documents relating to taxation, does the word actually appear in written texts. According to the qānūnnāme of Damascus (Barkan 220, Mantran and Sauvaget 5, Hammer, , Staatsverfassung, 1, 221Google Scholar), a synonym of dīmūs, in the sense of fixed money assessment, faṣl wa mafṣūl. This term occurs in Nuwaīrī (op. cit., viii, 260–1) and other Mamluk sources, where it has the same meaning. On the term dīmūs in modern Arabic see Dozy s.v. and Barthélemy s.v.

^{page 484 note 2} On the distinction between kāfirī olives, which pay a qasm of 1/2, and Islāmī olives, which pay a fixed money tax of 1 asper for 2 trees, see Notes and Documents, 19 and 41.

^{page 484 note 3} For further details on these taxes see Notes and Documents, 18 ff.

^{page 485 note 1} op. cit. ii, 266 ff. (above note 2, p. 474).

^{page 485 note 2} On the Jizye see further Notes and Documents, 2, 10 ff.

^{page 485 note 3} The rates are given in the qānūnnāme of Jerusalem (Barkan 219, Mantrau and Sauvaget 40–42).

^{page 486 note 1} The totals actually given in the registers contain several arithmetical errors, which have been corrected in the above tables. Thus, the Imperial ḵẖāṣṣ of Nabulus appears as 844,394, the governor's ḵẖāṣṣ and fiefs of Safed as 1,165,951, and the total of Nabulus as 1,353,107.

^{page 487 note 1} Budget figures of the ‘ Abbasid period vary in the neighbourhood of 300,000 dinars a year for the jund of Filasṭin and 100,000 for the jund of Urdunn, sometimes plus certain payments in kind (Jahsẖiyārī, , Kitāb al-Wuzarā ’, Cairo, 1938, 287Google Scholar; Ya'qūbī, , Kitāb al-Buldān, B.G.A. 7, 328–9, etc.).Google Scholar

^{page 488 note 1} See further Notes and Documents, 13 ff.

^{page 489 note 1} ibid., 16–17.

^{page 489 note 2} Qāqūn is one of the very few villages where the rate of qasm is changed during the period studied.

^{page 489 note 3} Probably an error for 5.

^{page 490 note 1} sic—but there are 15 names in the list.

^{page 491 note 1} cf. Notes and Documents, 18–19.

^{page 491 note 2} In the later registers, former qasm crops are included in the cash returns, indicating a restriction of the qasm system and an extension of the method of fixed money assessments.

^{page 492 note 1} In the registers of the Sanjaq of Tripoli, in the same period, wheat rises from 80 to 100 aspers per mekkūk. The gẖirāra of Damascus was 2 1/2 mekkūks of Tripoli, so that this is equivalent to a rise from 200 to 250 aspers per gẖirāra of 250 litres, as against 100 to 140 in the Sanjaq of Safed. On market prices for wheat and barley in Turkey see Güçer, Lütfi,‘ Le Commerce interieur des céréales dans l'Empire Ottoman pendant la seconde moitié du XVIème siècle’, Revue de la Faculté des Sciences Économiques de I'Université d'lstanbul, 2, 1949–50, 177Google Scholar ff., and Braudel, F., La Méditerranée et le monde méditerranéen à I'époque de Philīppe II, Paris, 1949,459.Google Scholar The Turkīsh and Venetian sources quoted by Güçer and Braudel show prices of 12 aspers a kaila and upwards for wheat in Constantinople. This is equivalent to about 300 aspers a gẖirāra. (The gẖirāra of Damascus was about 25 kailas of Constantinople—see Notes and Documents, 17 and 40.)

^{page 493 note 1} Interesting details on the processes of the manufacture of soap are given in the qānūnāame of Tripoli (Barkan 214–5, Mantran and Sauvaget, 69–70, where further references are given).

^{page 493 note 2} cf. Notes and Documents, 13, where reference to the second dye-house, included in the zi'āmet, was accidentally omitted.

^{page 493 note 3} On all these terms see Notes and Documents, 13–14 and 38–39.

^{page 494 note 1} cf. Notes and Documents, 19–20.

^{page 495 note 1} Where the proceeds of a tax are given as a round figure, it is safe to assume a muqāṭa'a, even if the register makes no mention of it. I have indicated the places where muqāṭa'a is specifically recorded.

^{page 495 note 2} This figure includes certain occasional revenues (bād-i havā) of the Sanjaq.

^{page 495 note 3} This figure includes the tax on the slave-market. This market is not separately mentioned, and the proceeds from it are presumably included under general headings. That the traffic in slaves was of some importance is shown by the customs and toll tariffs.

^{page 496 note 1} See Notes and Documents, 12– and 37–8.

^{page 497 note 1} The figure 63,000 in both registers includes the proceeds of the Qabbān, the Dār al-Wikāla, the vegetable market, and the market for thread and weaving.

^{page 498 note 1} For the reproduction and transcription of a similar table for the Sanjaq of Safed, see Eretz-Israel, Archaeological, Historical, and Geographical Studies, 2, Jerusalem, 1953, 195.Google Scholar

^{page 500 note 1} The arithmetic of this entry is unusually obscure.