¹ Neither the letter nor the treaty published here are mentioned by Bombaci, A., in his article ‘La collezione di documenti turchi del'lArchivio di Stato di Venezia’, Rivista degli Studi Orientals, XXIV, 1949, 95–107Google Scholar, but the treaty is listed in his ‘Regesti dei documenti turchi’ (in typescript at the Archivio di Stato) as number 58.

² Summaries of the contents of both the letter and the treaty are to be found in Predelli, R., I Libri Commemoriali della repubblica di Venezia, regesti, Venice, 1876–1914, XIX, nos. 148–9Google Scholar.

¹ Thus instead of tursilŪ; cf. the same verb in line 12 above.

² Instead of yajīū.

³ The tashdīd is probably due to contamination of tijar by the other plural form tujjar; besides one expects the nominative rather than accusative case here.

⁴ Yet another trace of the spoken language: naqdūhā instead of naqdīhā. I wish to thank Mr. David Cowan and Mr. Osman S. A. Isma'ūl for their observations on Arabic colloquial usage.

⁵ Headings (and marginal comments) are in the Libri Commemoriali otherwise usually in Latin. In this heading and in the one at the beginning of the treaty, both probably written by the scribe who entered the translations, certain abbreviations are used, such as Sig. for ‘signor’, Ill.ma for ‘Illustrissima’, and Capti for ‘capitoli’. For the ‘galee di barbaria’, see below, p. 469, n. 5.

⁶ ‘L'emir abuali’ is an insertion by the translator.

⁷ ‘Di quello’ is not a translation of ammā ba'du, which in Italian might be ‘dunque’, quindi’, ‘percio’, etc. Here ‘di quello’ appears to be a continuation of the preceding sentence ending in ‘operation’, although the Arabic demands a full stop at this point. See below, p. 465, n. 5.

⁸ Note ‘galie’ (sing. ‘galia’ or ‘galeay’) as a translation of the Arabic tarīda, for which the common Italian expression is ‘tarida’, derived of course from the Arabic. The term usually designates not a galley but a heavy transport or cargo ship (see Fahmy, A. M., Muslim seapower in the Eastern Mediterranean, London, 1950, 136–7)Google Scholar. The common Arabic words for galley are ghurāb and qit'a (Fahmy, op. cit., 133; and Wansbrough, , ‘A Mamluk letter of 877/1473’, BSOAS, XXIV, 2, 1961, p. 211Google Scholar, n. 8), though it appears that tarīda can also have this meaning (cf. the references to Quatremère, Mamlouks, Ia, p. 144, n. 18; p. 157, n. 33, in Fahmy, loc. cit.). I have retained galley in the English rendering, as these ships were the ‘galee di Barbaria’.

¹ The Alvise Pizzamano of Malipiero (see below, p. 469, n. 6), and the Alvixe Pizamano of Sanuto's Diarii (see below, p. 470, n. 1).

² ‘Mori’ (classical Mauros and Maurus, for inhabitants of Mauretania) is a common Venetian designation for the Arabic-speaking Muslims of North Africa and the Levant, though for the latter ‘Sarain’ is equally common.

³ Attenuated imperative form for ‘mandate’.

⁴ This insertion by the translator is puzzling. It appears to be a rendering of both meanings of wa'l-salām (‘That's all’ and ‘Greetings’).

⁵ Daif, rendered by the Italian translator ‘capo’, is employed in the Maghrib for ‘ruler’, see Dozy, Suppl., s.v. Other instances of this application of the term to Christian rulers can be found in Castries, de, Les sources inédiles de I'histoire du Maroc, Paris, 1905–Google Scholar, France I, 178, 183, n. 1, Spain I, 652, n. 1.

⁶ Leonardo Loredan (in the Arabic text ‘Lordano’; the name does not appear in the Italian translation) was Doge of Venice from 1501 to 1521. In his encyclopædia of chancery usage Qalqashandl lists dūj and dūk, but not duk as has our document, for the ruler of Venice (Subh al-a'shā, Cairo, 1920, VIII, 47); cf. al-dūj in the Mamluk document of 877'1473, BSOAS, XXIV, 2, 1961, 204

¹ For the ‘alāma see below, pp. 463–4. Neither of the two examples of it in our documents is clear (see plate i, line 3; plate II, line 4). Of the Amīr's names which have been transmitted in other sources, Mansūr b. Zayān al-Wattāsī, only Mansūr appears unequivocally in the 'alāma, as the first word, occupying the lower of two lines. Above it appears something like ywsf or ywsfā, followed only in Document 1 (plate I, line 3) by an unconnected yā, which might indicate a nisba. The name Yūsuf is an obvious conclusion, although it appears nowhere else among the Amīr's names nor does it quite account for the yā in the one example of the ‘alāma,. Nor, however, does the kunya Abū ‘Alī appear in other references to the Amīr so far discovered, though it occurs twice in our documents. We might thus tentatively propose the name: Abū ‘Alī Mansūr Yūsuf b. Zayān al-Wattāsī. Of three almost illegible Wattāsid ‘alāmāt de Castries (Sources, Spain I, 94) had deciphered one: that of Ahmad b. Muhammad al-Shaikh. For his method see the plate, loc. cit.

² The equivalent du‘ā’ in Mamluk documents is mahrūsa (cf. Amari, , I diplomi arabi nel R. Archivio Florentino, Florence, 1863, 198Google Scholar); in other North African documents cf. hamāha ‘llāh, kala'ahā 'llāh (Amari, op. cit., 4, 6, 76), and hamsahā 'llāh (Tisserant and Wiet, , ‘Une lettre de l'almohade Murtadâ au pape Innocent IV’, Hesépris, VI, 1, 1926, p. 32Google Scholar, n. 1).

³ The Italian ‘quella bona compagnia’ is an insertion but does not appear to alter the meaning of the original.

⁴ The two adjectives ādhiq and Kayyis have been omitted in the Italian text.

⁵ Fasl al-khflrīf signifies according to the Calendrier de Cordoue (edd. Dozy and Pellat, Leiden, 1961, 135 and 161) the period from the end of September to the beginning of November, a seasonal division probably also applicable to the near-by Mediterranean coast of Morocco. Fasl al-shitā’, winter, is the period from the second half of November to the beginning of March (Calendrier, 161 and 51).

⁶ ldhā sakhkharakum allāh is Qur'anic (e.g. xm, 2; xxn, 36; LXIX, 7; and cf. Lane, Lexicon, s.v. 2). It appears to have been unfamiliar to our translator who has rendered it ‘Se Dio vi doni aiuto’.

¹ The use of asīr for ‘slave’, and the verb asara for ‘to enslave’ (in art. III of the treaty, below; cf. the Italian text in both instances: ‘schiavo’ and ‘tegnir ne piar … per schiavo’) draws attention to the manner in which slaves were procured in the Maghrib (cf. also de Castries, Sources, France I, 181; though raqiq also occurs, cf. Amari, Diplomi, 4).

² See above, p. 449, n. 1 and 2.

³ Mansūr (al-Wattāsī), see above, p. 453, n. 1.

¹ See above, p. 451, n. 2, and the same word (written correctly) in line 11 below.

² For ‘Signoria’, i.e. the Republic of Venice. See below, p. 459, n. 8.

³ A mistake has been corrected by the scribe here. See also line 19 below, where the ordinal al-rābi‘ was inserted after the line had been written. In general the treaty appears to have been written less carefully than the letter, or at least more hastily, an impression given by sparser vocalization and the errors in the numbering of the articles, see below, p. 456, n. 5.

⁴ For ?

⁵ See above, p. 455, n. 3. There was clearly only sufficient place to insert the ordinal, without ‘aqd (‘article’).

⁶ The marginal numbers for articles V-VIII are in fact in the left-hand margin (see plate in), but for facility of reading are here placed in the right-hand margin.

¹ The marginal number was omitted here by the scribe.

² Thus consistently for dīnār, cf. lines 39 and 42 below.

³ Although the common Arabic expression signifying ‘tithe’ is ‘ushr, the fraction ‘tenth’ can also be ‘ushur, as it is here, just as ‘fourth’ is consistently vocalized rubu’ (see Wright, Grammar, third ed., I, 263 D).

⁴ The mark resembling kasra under fā may have been a slip of the scribe's pen.

⁵ The numbering, both marginal and in the text, for this article and the two following is incorrect. Appropriate emendations have been made in the English rendering below. The numbering in the Italian text is correct.

⁶ See above, n. 5.

⁷ Al-miray probably reflects ‘almiraglio’, see below, p. 461, n. 2.

⁸ See above, n. 5. The marginal article number has anyway been omitted by the scribe.

¹ The additional damma and the terminal jīm suggest that the scribe had begun to write another form of the verb. The vowel over jā has also been changed from fatha to damma.

² See below, p. 458, n. 7.

³ ‘Facti et opere’ as a rendering of qaulahu wa-'amalahu might better have been ‘parole et opere’, cf. paragraph II of the letter.

⁴ ‘In tempo de le galie’ is an insertion by the translator, probably, if he was Venetian, out of force of habit, as it is a common phrase in Venetian commercial records, and designates the period of trading in foreign ports.

⁵ For ‘pigliare’, to take or capture. For Venetian orthography see Boerio, M., Dizionario del dialetto veneziano, Venice, 1829Google Scholar.

⁶ The distinction between ‘terra’ and ‘paese’ is medieval (see Zingarelli, , Vocabolario delta lingua italiana, Bologna, 1958Google Scholar, 1592), and the former, modified by ‘marittima’ frequently designated a port. Elsewhere in our two documents madīna is translated not by ‘terra’ but by ‘cita’, and wa-'amaluhu by ‘o del suo paese’, or, as in art. I of the treaty, simply omitted.

¹ For ‘voglia’. This is a common Venetian spelling and, like ‘piar’ in art. II above, might appear in the most formal records of the Senate. Cf., however, ‘niun’ for ‘algun’ in arts. Ill, V, and X, an error which reflects the degree of literacy of our translator.

² ‘Angaria’ (or ‘angheria’, cf. Zingarelli, Vocabolario, s.v.), is a medieval term for impost, duty, or customs fee, for which the corresponding Arabic terms were many, e.g. ‘āda, darība, and mūjib (cf. Amari, Diplomi, 186, 188), and in this instance maghram (see below, p. 460, n. 2).

³ For ‘ducati cento’, cf. ‘per C°’ in the same article.

⁴ Arabic ra'īs; its appearance in this article together with the terms ‘patron’ and ‘Capitaneo’ is puzzling. The Arabic text only confuses the problem, see below, p. 461, n. 3.

⁵ ‘Se non el tornarano’ is an insertion by the translator.

⁶ ‘Sara mantegnudi’ is an insertion, but see below, p. 467.

⁷ ‘Etterzimi’ is not an accurate transcription of the scribe's nisba, which is al-Razīnī (see plate iv, line 50), and very likely indicates his membership of, or kinship with the Moroccan Berber tribe Banū Razīn (see Lévi-Provencal, , Histoire de I'Espagne musulmane, Leiden, 1950,I, 87)Google Scholar.

¹ See below, p. 466, n. 5.

² cf. Castries, de, Sources, France I, 8Google Scholar.

³ Castries, de, Sources, France I, 170Google Scholar, where the editor has translated al-anjad ‘le plus courageux’, which is more accurate than our Italian translator's ‘viotorioso’.

⁴ See above, p. 453, n. 1.

⁵ cf. madīnat luklca wa-'amaluhu in reference to Lucca and its vicinity, in a Pisan treaty with the Banū Ghanīya in 1184 (Amari, Diplomi, 232).

⁶ See below, p. 466, n. 5.

⁷ The Italian translator has been fairly consistent in the rendering of names for ships: in paragraphs III, IV, VI of the letter and articles VIII, IX, X of the treaty tarīda is translated ‘galia’ (see above, p. 451, n. 8); in articles II and V of the treaty jafn is rendered ' navilio ‘which, however, is used in article VI to translate safina. Like ‘narilio’ both jafn and safina are more or less generic names for ship, and the two Arabic terms seem to occur interchangeably in North African documents (see Brunschvig, , La Berberie orientals sous les Hafsides, Paris, 1940–7, II, 98Google Scholar; and de Castries, , Sources, France I, 9, 180Google Scholar). Safīna can mean specifically a transport or cargo ship (see Fahmy, Muslim, sea-power, 162) and might better have been translated by a Venetian term with the same application, such as ‘arsilio’ (see Latrie, Mas, Traités de paix et de commerce et documents divers concernant les relations des chre'tiens avec les arabes de I'Afrique septentrionale au moyen âge, Paris, 1866, p. 267Google Scholar, n. 1). Occasionally, as in the Almohad-Pisan treaty of 1186, a distinction is made between Muslim and Christian ships, the former being designated asātīl (sing, ustūl, see Fahmy, Muslim sea-power, 127) and the latter marhab (Amari, Diplomi, 21, 22).

⁸ The transcription of European words in the Arabic texts of medieval commercial treaties was not uncommon, especially for proper nouns, such as al-shunyūrīya for ‘Signoria’ here (cf. kumūn bīsh for ‘Comune di Pisa’ in the Hafsid-Pisan treaty of 1313, in Amari, Diplomi, ). Rendering ‘s’ with shīn appears, moreover, to have been common in the Maghrib and Andalus (cf. Ishbanīya for ‘Hispania’, and even Andalush in Lévi-Provencal, , La Péninsule Ibérique au moyen âge, Leiden, 1938, 4Google Scholar, 8).

⁹ The reference here is of course to amān or ‘safe-conduct’. The passive participle mu'amman is employed, and in article V a noun of place: ma'man, which appears to signify ‘position’ of ‘status of safe-conduct’. For other forms of the verb, all of which mean that amān has been granted, cf. Amari, Diplomi, 154, 227. See also El, second ed., s.v.; Brunschvig, , La Berberie, I, 430–40Google Scholar; and Heffening, W., Das islamische Fremdenrecht, Hannover, 1925, 88–96Google Scholar.

¹ But see article X, below. A clause prohibiting the practice of collective responsibility was common in the commercial treaties concluded between medieval Muslim and Christian rulers, though not without certain qualifications, as in article X of our treaty, and in various of the Hafsid treaties concluded with Venice, Genoa, and Pisa (cf. for example, Mas Latrie, Traités, Docs. 89, 120, 197, 200, 204), where the consul was held responsible for the debts of compatriots who departed owing Muslims money (cf. also Brunschvig, , La Berberie, I, 436)Google Scholar.

² Maghram (plur. maghǎrim) is a common term in North African documents for impost and customs duty (cf. Amari, Diplomi, 134, 148, 162, 174); maghram may also designate an impost on the property of a tribe (cf. Brunschvig, , La Berberie, II, 195)Google Scholar.

³ See above, p. 456, n. 1. From here on the numbering refers to the corrected roman numerals accompanying the original text.

⁴ Here, and in article IX below, sila’ (sing, sil'a) is used for ‘goods’ or ‘merchandise’, as is common in North African documents (cf. Amari, Diplomi, 17–22 passim), while the common terms in Mamluk documents are badā'ī’ and matājir. In the Maghrib sil'a can also be used in the sense of shoddy wares, or simply junk (cf. Dozy, Suppl., s.v.). In our document the Italian translation of sila‘ is ‘roba’, which is slightly misleading as it is employed in articles II and VI to translate māl (and its plural amwāl), which signify ‘property’ in the sense of ‘possessions’ (and sometimes specifically ‘money’), in contrast to ‘merchandise’. A better Italian translation of sila‘ might have been ‘merci’ or ‘mercadantie’.

⁵ This seems a curious way to express ‘seven and one half’, apparently even to the scribe, for he explains it at once. The Italian translator did not bother to include the calculation. I would only tentatively suggest that the originator felt obliged to explain his rate of customs duty in terms of a tenth or ‘tithe’ (‘ushr or ‘ushur, cf. above, p. 456, n. 3) as that was according to Muslim practice the traditional way of expressing customs duties in sharī'a terminology (cf. Heffening, , Fremdenrecht, 127–9)Google Scholar.

¹ ‘Aiā mā dhukira has been omitted in the Italian translation.

² The word in the Arabic text is al-miray, which appears to provide yet another link in the long chain of etymologies proposed for the word ‘admiral’. At least three of these, all signifying a ‘commander of naval forces’ and discovered in Arabic sources, are similar to ours: al-mirānt and al-mīland (in Ibn Khaldūn's history of the Berbers, cited Brunschvig, , La Berberie, n, p. 94Google Scholar, n. 3), and al-mirālyā (in a Mamhik letter to the King of Aragon in 703/1304, in M. Alarcón y Santón and Ramón De Linares, Garcīa, Los documentos arabes diplomāticos del Archivo de la Corona de Aragón, Madrid-Granada, 1940, 350Google Scholar). For related etymologies see also Lokotsch, , Etymohgisches Wōrterbuch der europäischen Wōrter orientalischen Ursprungs, Heidelberg, 1927Google Scholar, no. 69, and Fahmy, , Muslim sea-power, p. 139Google Scholar, n. 2. A difficulty in our document is the intended distinction between al-miray and al-qabtǎn. The latter refers to Alvise Pizamano, while the former is translated in the Italian text as ‘patron’. Although ‘patron’ could be ‘owner’, it is not unlikely that this title also refers to Pizamano, and that it was he who was awarded the 15 dinar exemption from customs duty.

³ The grammatically puzzling wa-hūwa ra'īsunfī kulli tarīdatin has been omitted in the Italian text. I would suggest tentatively that it is a circumstantial clause and meanssomething like ‘while (or so long as) he is ra'is inhis (respective) galley’. It is not clear how many galleys there were and thus how many officers of this rank were granted the exemption. According to the Italian translation (‘Et i rais de l'altra galia’) there can have been a total of only two galleys, though the ‘galee di Barbaria’ sometimes included three (see Latrie, Mas, Traites, Docs., 269Google Scholar).

⁴ According to article VII above, the Muslim buyer was required to pay a duty of seven and one half per cent, and the Venetian seller compelled to pay that amount when the buyer defaulted. In article X, however, the exemptions appear to be based upon the presupposition that the seller, too, pays a duty of seven and one half per cent. While far from clear, it may be that the total duty levied by the Amīr was fifteen per cent, one half payable by each party to a transaction, and that in article VII the amount which the Venetian was required to pay on behalf of the Muslim was in addition to his own obligation With respect to the amount of the exemption the Arabic and Italian texts do not appear to agree. In the former the miray is granted a remission of 15 dīnārs, and each ra'īs (apparently) one of seven and one half dīnārs. In the Italian translation the provision for the miray is the same as in the Arabic, but for each ra'īs the total amount payable is seven and one half dīnārs, a far more generous arrangement than he had been given in the Arabic text. The difference would depend of course upon how much the ra'īs was able to sell, a contingency to which the otherwise puzzling insertion in the Italian text, ‘debi esser tenutoconto’, may refer.

⁵ Yutliqūhu has been omitted in the Italian translation.

⁶ See above, article IV, and p. 460, n. 1.

¹ For these characteristics in other North African documents see Tisserant and Wiet, , ‘Une lettre’, Hespéris, VI, 1, 1926, 29Google Scholar; and cf. Wright, Grammar, third ed., I, 14 D.

² See Amari, Diplomi; Mas Latrie, Traiteés; de Castries, Les sources inedites (first series, volumes for France and Spain); Lévi-Provencal, Trente-sept lettres officielles almohades, Rabat, 1941 (translation and commentary in Hespéris, XXVIII, 1941); Alarcón and Garcīa, Documentos; and Brunschvig, La Berbérie, bibliography.

³ The classical pages of Qalqashandī on the form of MaghribI documents (Subfi al-a‘sha, vn, 78 ff.) are translated in Tisserant and Wiet, ‘Une lettre’, Hesperis, VI, 1, 1926, 27–8, though it must be remarked that our documents conform in practically no respect with the precepts set out there. On the diplomatic form employed in the Almohad chancery, see Lévi-Provencal, , ‘Un recueil de lettres officielles almohades’, Hespéris, XXVIII, 1941, 1–80Google Scholar, especially 10–19. Observations of a very general nature on North African diplomatic are to be found in the article of G. Colin in El, second ed. (‘Diplomatic’), and in more detail in de Castries, ‘Les signes de validation des chérifs saadiens’, Hespéris, I, 3, 1921, 231–52Google Scholar.

⁴ cf. de Castries, ‘Signes’, Hespéris, I, 3, 1921, plate I; Amari, Diplomi, 17.

⁵ cf. de Castries, Sources, France 1, 178, 347, 359, 383; Spain I, 92, 126, and ‘Signes’, Hespéris, I, 3, 1921, plate IX. See also the inscriptions in Colin and Mercier, Corpus, II, 31, 81.

⁶ For the Ottoman tughrā see Wittek, P., ‘Notes sur la tughra ottomane’, Byzantion, XVIII, 1948, 311–34Google Scholar; xx, 1950, 267–93; and for the Mamluk ‘alārm, Wittek, Ibid., xx, 290–1. It is very likely that the introduction of the ruler's name into the sign manual was an eastern innovation, see Cahen, , ‘tuñrā seljukide’, Journal Asiatique, CCXXXIV, 1943–5, 169–70Google Scholar, and Wittek, Ibid., xx, 288.

¹ The hamdala was apparently an Almohad innovation, see Colin, in El, second ed. (‘Diplomatic’), Lévi-Provencal, , ‘Un recueil’, Hespéris, XXVIII, 1941, 17–19Google Scholar, Tisserant and Wiet, , ‘Une lettre’, Hespéris, VI, 1, 1926, 17–19Google Scholar, vi, 1, 1926, plate I, and de Castries, ‘Signes’, Hespéris, I, 3, 1921, 236–8. De Castries's excellent study applies almost exclusively to the Sa‘dian chancery, subsequent to the period of Wattasid rule in Morocco. The Hafsids, contemporaries of the Wattāsids, did include their names in the ‘al'ma, (see Brunschvig, , La Berbérie, II, 62–4Google Scholar).

² Three examples of Wattāsid ‘aldmāt, each containing only the name of the originator/ruler, appear in de Castries, ‘Signes’, Hespéris, I, 3, 1921, plate II (for the complete documents, cf. de Castries, Sources, Spain I, 94, 142–3, 426–7, plates I, III, VI).

³ cf. the position of the three examples in n. 2, and de Castries's remarks on another form of the sign manual, the sahīha dhālika, in ‘Signes’, Hespéris, I, 3, 1921, 234 and plate I. The Sa'dian ‘alāmx0101;t appear to have been at the top of the document. Our documents contain no corroboratio.

⁴ cf. Colin, El, second ed. (‘Diplomatic’). Ibn Khaldūn held this position for a time in the Hafsid chancery at Tunis.

⁵ The larger script is usually though not invariably indication of an ‘alāma. Amari's note in reference to a Marīnid-Pisan commercial treaty of 1358 (Diplomi, App. 1–8, and p. 69, n. 16) requires modification. The date (kutiba fī) was frequently written in thuluth (according to Qalqaahandī, cited Tisserant and Wiet, , ‘Une lettre’, Hespéris, VI, 1, 1926, p. 28Google Scholar), and was not to my knowledge employed as an ‘alāma.

⁶ In the Venetian chancery there were rules for determining precedence between the Doge and his many correspondents (see Latrie, Mas, ‘Listes des princes et seigneurs’, Bibliothéque de l' école des chartes, 1864–5Google Scholar), as there were also in the Mamluk chancery (Qalqaahandī, Subb al-āsha, VIII, 25–53). It is noteworthy to find that translators occasionally changed the order of inscriptio and intitulatio as a compliment to the ruler by whom they were employed (see Tisserant and Wiet, , ‘Une lettre’, Hespéris, VI, 1, 1926, p. 52Google Scholar, n. 1, and Amari, Diplomi, 1, 255, in a commercial treaty between Pisa and the Banū Khurasan in 1157). Thus is probably to be explained the inversion in a translation of a Mamluk letter to the Doge in 877/1473, for which I had offered another explanation (see ‘A Mamluk letter’, BSOAS, XXIV, 2, 1961, p. 204Google Scholar, n. 1).

¹ See above, p. 452, n. 5.

² cf. also the Qur'anic (xx, 47) expression wa'l-salām ‘aiā man ittabaāa 'l-hudā (Hamidullah, M., Corpus des traités et lettres diplomatiques de I'lslam, Paris, 1935, 18Google Scholar; and Castries, de, Sources, France I, 168Google Scholar), or the phrase anāra 'llāhu ta'ālā baṣīratahu bi-taufīqihi wa-irshādihi (addressed to Pope Innocent IV by the Almohad caliph Murtaḍa, cf. Tisserant and Wiet, ‘Une lettre’, Hespéris, vi, 1, 1926, 30). See also Amari, Diplomi, 1: ahdā 'llāhu taufīqahum. The formula commonly used in the Ottoman chancery for Christian rulers was khutimat ‘aivdqibuhu bi'l-khair (Kraelitz, Osmanische Urkunden in türhischer Sprache, Wien, 1921, 24), and in the Mamluk chancery aṭāla ‘lldhu baqd'ahu (Qalqashandī, Ṣubḥ al-a'shā, VIII, 35; Alarcón and García, Documentos, 344, 350, 355, 360). It might be that all such ‘good wishes’ contain at least implicitly a reference to a conversion to Islam (cf., for example, the Venetian translation of wa-allāh al-muwaffiq bimannihi wa-haramihi in the Mamluk document of 877/1473, BSOAS, xxiv, 2, 1961, 202–3).

³ See above, p. 464, n. 6.

⁴ A more common du‘ā’ for a ruler was simply aslaha 'llāhu ḥālahu (cf. de Castries, Sources, Spain i, 132, 143, 235, 285, 365, 427, 463). The Amīr's titles in our documents (see plate i, line 3, and plate n, lines 3–4) are, in comparison with those used by the Mamluk and Ḥafṣid rulers, remarkably modest. The expression in the letter al-rājī raḥnmta rabbihi appears to have been common, for Orient as well as Occident, in funereal inscriptions, cf. Colin and Mercier, Corpus, II, 243—4 (tomb of a Muslim doctor), and M. van Berchem, Matériaux pour un corpus inscriptionum Arabicorum, Cairo, 1894–1930, Egypt i, pt. 1, pp. 157, 277, 428 (tombs of Mamluk amīrs).

⁵ This formula, which means ‘and now to the matter at hand’, was regularly employed in the Arabic chanceries of Egypt; see Björkman, Beiträge zur Geschichte der Staatslcanzlei im islamischen Ägypten, Hamburg, 1928, 13, 116. For its use in Maghribī documents, cf. de Castries, Sources, France I, 9, 359; Lévi-Provencal, ‘Un recueil’, Hespéris, xxviii, 1941, 13, 16 (where in Almohad protocol it introduces the ḥamdala); Amari, Diplomi, 1, 36, 78.

¹ cf. ‘A Mamluk letter’, BS0AS, xxiv, 2, 1961, 210–11Google Scholar.

² cf. Björkman, StaatskanzUi, 28, 53; and for the Fāṭimid chancery, S. Goitein, ‘Petitions to the Fātimid caliphs from the Cairo Geniza’, Jewish Quarterly Beview, XLV, 1, 1954, 30–8; and S. M. Stern, ‘A Fāṭimid decree of the year 524/1130’, BSOAS, xxiii, 3, I960, 448–9.

³ See Amari, Diplomi, 185.

⁴ cf. Amari, Diplomi, 215; and ‘A Mamluk letter’, BSOAS, xxiv, 2, 1961, 206, paras. V, VI.

⁵ cf. Amari, Diplomi, 4, 15, and App., 1–8. The term ‘aqd also designates ‘article’, and mai'āqada the collective articles, or ‘treaty‘9 (cf. Amari, Diplomi, 2, 4, 6, 9, 230, 231), as in fact they do in our documents (see plate i, lines 10–11; plate II, lines 1, 7, and thereafter for each of the ten articles, concluding with the same phrase on plate iv, line 47). Where mu'āqada appears in construct with (plate i, lines 10–11) I have translated it ‘conclusion of articles’, but where it appears alone, simply as ‘treaty’. Cf. the phrase al-muhādana al-ma'qūda (in a commercial treaty between Pisa and the Banū in 1184), translated by Amari ‘tregua convenuta’ (Diplomi, 235). The corresponding Venetian terms are ‘capitoli’, ‘pacti’, and ‘acordi’, used without much consistency in the translations here. Another common Venetian expression is ‘comandamenti’, not used in the Italian version of our documents, but almost exclusively for the treaties granted Venice by the Mamluk sultans, called in Arabic marāsīm, or ‘decrees’ (cf. ‘A Mamluk letter’, BSOAS, xxiv, 2, 1961, p. 211, n. 4).

⁶ It was the custom of the Venetian Signoria to send detailed instructions, amounting in fact to a petitio, arranged in paragraphs, with envoys dispatched to foreign rulers for the purpose of securing commercial privileges. These instructions, in Latin or Italian, became the basis for the wording of the articles. The Republic of Florence adopted this practice from Venice on the occasion of her first embassy to the Mamluk Sultan in 1422 and again in 1489 (cf. Amari, Diplomi, 331–5, 374–81).

¹ In our treaty, the first article is a confirmation of peace (ṣulḥ) between Bādis and Venice, while the second article refers specifically to amān.

² For the Mamluk chancery, see Ernst, Die mamlukischen Sultansurkunden des Sinai-Klosters, Wiesbaden, 1960, xxviii-xxxi; and for the Ottoman chancery Kraelitz, Osmanische Urkunden, 27ā9.

³ For a similar expression, without the imperative, in a Sa'dian document of 1559, cf. de Castries, Sources, France i, 172. Cf. fa-yuḥīṭ 'ilman bi- in a Mamluk document (‘A Mamluk letter’, BSOAS, xxiv, 2,1961, 208), and the Ottoman bilesiz (Kraelitz, Osmanische Urkunden, 128).

⁴ See Björkman, Staatskanzlei, 116–17; Ernst, Sultansurkunden, xxxvi-xxxvii; ‘A Mamluk letter’, BSOAS, xxiv, 2, 1961, 210, 213, n. 4–7.

⁵ See Björkman, Staatskanzlei, 116; Lévi-Provençal, ‘Un recueil’, Hespéri s, xxviii, 1941, 13, 14, 16; Amari, Diplomi, 6, 37, 76–7.

⁶ See Bresslau; Handbuch der Urkundenlehre, third ed., Berlin, 1958, I, 74, 77.

⁷ cf. Amari, Diplomi, 37, 77, and for bi-ta'rikh, the Mamluk-Venetian treaty of 1507 (ASV, Miscellanea di Atti Diplomatici e Privati, doc. 1576). In our documents bi-ta'rīkh appears in the letter, while in the treaty the date is simply appended to the sentence wa kataba … Muḥammad b. Aḥmad (see plate I, line 21, and plate rv, lines 49–53).

⁸ cf. Lévi-Provencal, ‘Un recueil’, Hesp'ris, XXVIII, 1941, 16; and Amari, Diplomi, 37, 77.

⁹ cf. Amari, Diplomi, 22; Colin and Mercier, Corpus, n, 13, 77, 78, 90; de Castries, Sources, France I, 360. Not used by the Mamluk chancery, these expressions are common in Ottoman documents (cf. Kraelitz, Osmanische Urkunden, 34–8).

¹ cf. de Castries, Sources, France I, 9, 173, 360; Colin and Mercier, Corpus, I, 5, 43, 53, 59; the use of sana appears to have spread after the Ottoman conquest of portions of North Africa in the sixteenth century, and there are even examples of ‘ām and sana used in the same document (cf. Colin and Mercier, Corpus, IT, 88).

² cf. the Pisan-Banū treaty of 1184, and the Pisan-Marīnid treaty of 1358 (Amari, Diplomi, 235, App., 8).

³ Unlike Ottoman documents, where it is generally included (see Kraelitz, Osmanische Urkuvden, 34–8).

⁴ See Colin in El, second ed. (‘Diplomatic’), de Castries, ‘Signes’, Hespéris, i, 3, 1921, 234; Lévi-Provençal, ‘Un recueil’, Hespéris, XXVIII, 1941, p. 18, n. 35, and Rosenthal, ‘The technique and approach of Muslim scholarship’, Analecta Orientalia, xxiv, 1947, 39. But cf. the reservations expressed in Tisserant and Wiet, ‘TJne lettre’, Hespéris, vi, 1, 1926, p. 28, n. 1. In these examples the tail of the hā is curved towards the right, as in a yā mardūd.

⁵ By the treaties of Tordesillas (1494) and Villafranca de Xira (1509); Bādis was called by the Spaniards Velez de la Gomera. See Cour, La dynastie marocaine des Beni Waṭṭās, Constantine, 1920, 125; Terrasse, Histoire du Maroc, Casablanca, 1950, II, 112, 116; and El, second ed. (‘Bādis’).

⁶ For the approximate date of Muhammad al- death, cf. de Castries, Source s, Spain I, plate iv (a genealogical table of the Waṭṭāsids). According to Cour, Beni Waṭṭās, 92, the date was 910/1504; and in Zambaur, Manuel de généalogie, 80, he is not listed as a ruler, but as the brother of the ruler Yaṭyā.

¹ See Cour, Beni Waṭṭas, 92–112, esp. 93, 94, 98; Terrasse, Maroc, n, 148; Zambaur, Manuel, 80.

² Cour, Beni Waṭṭās, p. 125, n. 2, where the author cites three sources, one Arabic and two Portuguese, for this information. Shortly afterwards, in 1526, Bādis achieved a kind of fame as the fief of the Waṭṭāsid pretender Abū Ḥassun ‘Alī b. Muḥammad al- who because of his grant bore the nisba al-Bādisī (Cour, Beni Waṭṭās, 154–204; de Castries, Sources, indexes for France I, n, Spain I, n; and Terrasse, Maroc, n, 143, 167, 170, 181, 260).

³ See Terrasse, Maroc, n, 122, 133; and Julien, Histoire de L'Afrique du Nord, Paris, 1956, II, 250–5.

⁴ Leo Africanus, Description de l'Afrique (ed. A. Epaulard), Paris, 1956. His observations on Bādis are few but important (cf. 274–6, 277, 279, 292, 294, 295) as his sojourn in Morocco coincided almost exactly with the date of our documents. The citadel in Bādis is said to have been constructed by the Marīnid Abū Yūsuf Ya'qūb b. 'Abd al-Ḥaqq (656–85/1258–86), see Terrasse, Maroc, II, 78.

⁵ The ‘galee del trafego’ were established in 1432 and plied between Venice, the Maghrib, and the Levant. The ‘galee di Barbaria’, established in 1440, were restricted to the coasts of North Africa, Spain, and Sicily (that is, Syracuse, Tripoli, Jerba, Tunis, Bougie, Algiers, Oran or Bādis, Honain or some near-by port, to be decided by the Captain and his consultative body composed of merchants and called ‘Consiglio di Dodici’, Malaga and/or Almeria, and on the return to Venice, second calls at Tunis and Syracuse), cf. Mas Latrie, Traités, Docs., 266–9 (a commission for the captain of the ‘galee di Barbaria’ for 1508–9, which despite the author's supposition to the contrary, cf. p. 266, n. 1, and 259, did sail, under the command of Bortolo da Mosto, not Piero Mulla (Sanuto, Diarii, Venice, 1877–1900, vi, 603)). See also Cour, Beni Waṭṭās, 99; and Terrasse, Maroc, n, 155, where the Venetians are mentioned as having concentrated their activities at Badis, and the Genoese at Alcudia.

⁶ Malipiero, Annali Veneti (Archivio Storico Italiano), Florence, 1843–4, 635. Pizamano is here spelled ‘Pizzamano’.

¹ Sanuto, Diarii, vi, 29, 105, 226, 415, 468, 489.

² For a detailed discussion of the legal and commercial contents of commercial treaties concluded between Muslim and Christian powers in the Middle Ages, see Mas Latrie, Traités, 83–116, 185–224.

¹ The slave trade was intimately related to the activity of corsairs along the Moroccan coast, for which see Cour, Beni Waṭṭās, 99–100; Mas Latrie, Traités, 236–7, 260–1.

² For example, a treaty between Pisa and the Banū in 1184 for ten years and six months (Amari, Diplomi, 235); between Florence and the Ḥafsids in 1445 for 31 years (Amari, Diplomi, 171); between Portugal and the Waṭṭāasids in 1538 for 11 years (de Castries, Sources, Spain i, 83–4).

³ See Mas Latrie, Traités, 184, 330–1, Docs., 257, and ASV, in the archive entitled ‘Terminazioni ed incanti di galere’, extracts of which are published in Archivio Veneto, Serie 5, n, 250 ff.

⁴ A possible exception is grain, see Brunschvig, La Berbérie, II, 266, 438, for Venetian imports from Tunis, which were also largely agricultural products.

⁵ See Mas Latrie, Traités, Docs. 273–5.

I should like to thank Professors Paul Wittek and Bernard Lewis for their helpful suggestions in the preparation of this study, and at the same time to express my gratitude to the director and staff of the Archivio di Stato in Venice for their kindness during the period of my researches there in 1960.