^{page 247 note 1} This Persian name, which appears to-day on the maps as Qais, is very differently spelled in Arabic and European sources, cf. the detailed analysis in M. Streck's article in the Encyclopedia of Islam, s.v. Ḳais. In the documents dealing with the events described here, the island is invariably called Kīsh. However, in the literary sources recounting it, the name is spelled qys in the narrative and ks in the quotation of the words of the sailors coming of that island, which change is obviously intended to show them as speaking Persian, q in qys was no doubt pronounced in the Bedouin-South Arabian way as a g, formed very far back, while k in ks also represents a g, the Persian g being written as a k with three dots or a stroke superscript, these diacritics, however, being often omitted. Thus the actual name of the island probably was ‘ Guess ’, and so indeed it is spelled in some of the accounts of the European travellers who first visited it. See. Streck, s.v. A full discussion of the medieval sources on this island is found in Wilson, Arnold T., The Persian Gulf, 1928, pp. 95–100.Google Scholar

^{page 248 note 1} Kay, H. C., Yaman, its Early Medieval History, London, 1892, p. 79Google Scholar, of the English translation. Mujāwir, Ibn, ed. Loefgren, , 1951, p. 123, 1. 19,Google Scholar obviously refers to one of Bilāl's sons.

^{page 248 note 2} Kay, op. cit., p. 80.

^{page 248 note 3} At that time one could live in Cairo for a whole month for one dīnār.

^{page 248 note 4} Kay, op. cit., p. 66; Encyclopedia of Islam, Karam, s.v., vol. 2, p. 795.Google Scholar Other details about the tributes paid by the governors of Aden are to be found in Loefgren, , Texte, p. 65,1. 5 sq.Google Scholar

^{page 248 note 5} This is at least what Idrisi, the famous 12th-century traveller, reports, cf. Wilson, op. cit., p. 98.

^{page 248 note 6} Description of Yaman, ed. Loefgren, O., Leiden, 1951, pp. 124, 1. 5—125, 1. 8Google Scholar, ib., 1. 15—126, 1. 2. This passage is contained also in the same editor's Arabische Texte zur Kenntnis der Stadt Aden im Mittelalter, 1936, pp. 43–5.Google Scholar It may be alluded to or discussed by Hunter, F. M., An Account of the British Settlement of Aden, 1879Google Scholar, or Ferrand, G., Journal Asiatique, ser.. II, torn. 13 (1919), 472–483Google Scholar, Loefgren, cf., Texte, p. 19, note 3.Google Scholar However, owing to the well-known present conditions, I am unable to consult these books, which have remained on Mount Scopus. For the same reason I have to be excused for not using Ibn al-Jauzī's Muntazam and other sources on general history, which may contain details about the siege of Aden—although this is not very likely.

^{page 248 note 7} The date itself must be erroneous. For Saba’ the Zurai‘id, for whom the castle al-Akhḍar was stormed and taken by his freedman Bilāl b. Jarīr, died in 533 (1138).

^{page 249 note 1} So far I have collected 32 items relating to this successful Indian merchant, who was also a public-minded man and a poet (in Hebrew, of course, although his correspondence, as far as found so far, is almost exclusively in Arabic). In the orbit of Muslim civiḷization there is, of course, nothing surprising in finding a great businessman indulging in writing poetry. As a Muslim counterpart ‘ the poet Takrītī ’, an Iraqian merchant, who had lost everything in shipwreck, but worked his way up in Aden with the aid of his poetical talent, could be adduced, Makhrama, Abu cf.'s History of Aden, ed. Loefgren, , vol. 2, pp. 22–38.Google Scholar

^{page 249 note 2} Many references to the price of paper are found in the Genizah letters.

^{page 249 note 3} e.g. glass cups preserved in the typically Yemenite baskets, called shutūt (sg. shatt, in common use to-day), mats from Berbera, in East Africa, a leather carpet (as ‘ table-cloth’, or rather as ‘ table ’), an iron pan (in other years he ordered a pan made of stone—in use in Yemen up to the present day), a sieve, soap, and a quantity of cloth, mostly of Egyptian manufacture.

^{page 250 note 1} Numbers HN1 and HN7 of my collection.

^{page 250 note 2} As may be learnt from HN3, dispatched from Aden in the autumn of 1137.

^{page 250 note 3} No. G13.

^{page 250 note 4} Kay, op. cit., p. 67. cf. Ibn Mujawir, p. 122,11. 6 and 8, p. 124,1. 4.

^{page 250 note 5} ‘ Zurai‘ids ’ is used here a fortiori, Zurai‘ being the grandfather of Saba’. This is already the usage of the medieval Muslim historians, e.g. Ibn al-Mujāwir, p. 124, 1. 8, calls the overlords of the castles al-Khaḍra’ and Ta’ker‘ The sons of Zurai‘ ’, although only one of them actually was a descendant of Zurai‘.

^{page 251 note 1} Loefgren, cf., Texte, 1, p. 65,1. 4.Google Scholar To-day Ṣīra is the name of the small island which protects the outer, eastern, port of Aden, as may be seen on any map of the town.

^{page 251 note 2} i.e. he would be prepared to serve them, just as he held the castle at present for his Sulaiḥid overlord.

^{page 251 note 3} In the text: jāshū, a Persian word designating sailors, still used in the Persian gulf, Loefgren, cf., Texte, 1, p. 44, note 3, and ii, p. 25.Google Scholar Ibn Mujāwir, indeed, reports in his description of Qais (Kīsh) that its prince had neither cavalry nor infantry, all the people of the island being mariners, cf. op. cit., p. 100.

^{page 251 note 4} Loefgren, cf., Texte, p. 45, note 11.Google Scholar There could hardly be a doubt that the reading was correct. The famous Dair al-Jamājim,‘ The Convent of the Skulls ’, in Iraq, the scene of a great battle in A.D. 701, would be a parallel; cf. Encyclopedia of Islam, s.v.

^{page 252 note 1} The manuscript clearly has nrmtyn; however, I have the impression that the very cursively written MS. is only a copy of the original, the copyist having mistaken b for n. Owing to the great risks, usually two or even three copies of the same letter were sent in different ships. One writer mentions that he had sent a letter through three different persons by three different routes and each in two copies, one in Hebrew and one in Arabic. Of five letters sent to India I have actually found two copies. Document 2 also may be a copy, being written in one instead of two words in 1. 17, which is a typical copyist's error.

^{page 252 note 2} Burma is listed by the 10th-century Muslim geographer Muqaddasi, BOA., 3, 1906, p. 32,1. 2Google Scholar, and discussed by de Goeje in BOA., 4, pp. 188 and 231Google Scholar, which, owing to the circumstances referred to in note 3, p. 3, I am unable to use. For the same reason I could not see Kindermann, H.'s ‘ Schiff’ im Arabischen, 1934.Google ScholarFahmy, A. M., Muslim Sea-power in the Eastern Mediterranean, 1950, p. 155Google Scholar, says that no information about it could be found, while Hourani, G. B., Arab Seafaring in the Indian Ocean, pp. 98 sq.Google Scholar, if I am not mistaken, does not mention it at all.

^{page 252 note 3} At first I was inclined to explain naranjiyāt ‘ ships formed like oranges ’; in this case, however, Ibn Mujāwir would not have said abrām an-naranjiāt, but simply an-naranjiyāt. Loefgren, , Texte, vol. 2, Glossar, p. 59Google Scholar, explains the term by the superstitious custom of sailors passing Sokotra of putting a pot (the text has qidr) with various kinds of food into the sea.

^{page 252 note 4} Reproduced in Wilson, op. cit., p. 98.

^{page 252 note 5} Dr. Serjeant kindty consulted for me Kindermann's ‘ Schiff’ im Arabischen, where indeed Hindustani ‘ majra ’ or ‘ mashwa ’ (Miles, cf., JBAS., 21 (1889), pp. 662 and 821Google Scholar) are given. Kindermann contains no reference to either shaffāra or jāshujīya.

^{page 253 note 1} Thus in one document, Yll, one merchant alone had in a shaffārah fourteen buhārs (a buhar contained, in the letters referred to, 300 rotl or pounds).

^{page 253 note 2} cf. the Arabic ‘ down the stream’, which corresponds to the Aramæic shippūla known from the Talmud, cf. Fraenkel, 230.

^{page 254 note 1} In a letter, written in Tarim, Hadhramaut, on 9th December, 1953, Dr. R. B. Serjeant informs me that the term dīwānī is indeed found in a Yemenite manuscript dealing with tribal law in the sense of ‘ professional soldier ’.