1 See my article: The Epistle of the Fatimid Caliph al-Amir (al-Hidkya al-Amiriyya); its date and its purpose, in: J.R.A.S., ig5o, p. zo ff. ' Cf. G. Wiet in: Journal Asiatique, 1921, II, p. 112, for the different dates mentioned by the historians. (Add to the authorities mentioned there, e.g., Ibn Hammad, ed. Vonderheyden, p. 60; al-Warraq, in: al-Baydn al-ll�ughrib, ed. Colin and Levi-Provencal, p. 311; al-Bustkn al-idmi'

cf. below, p. 198). The 'i7yMK al-akhbar (VII, 313, in the manuscript at my disposal), like the Tayyibi tradition in general, gives the year as 526; it is difficult to say how the mistake arose. As to the day, the '!7yt<)t has Tuesday the 3rd Dhu-I-qa'da, as have, e.g., Ibn Khallikan, the Mir'ai al-zaman and al-Qalqashandi. 1 So the various articles of H. F. Hamdani: The Life and Tinzes of Queen Sayyida Arwa the Sulaytiid of the Yemen in: Journal of the Central Asian Society, 1931, p. 505 ff.; The History of the Isma'ili Da`xewat and its Literature during the last phase of the Tatimid EmPire, J.R.A.S., 1932, p. 126 ff.; Some unknown Ismd'ili authors and their works, J.R.A.s., 1933, p. 359 fi- Their merit lies precisely in acquainting Western scholars for the first time with the outlines of Ismaili history as it appears in the light of the literature of the Bohra. 2 So the article Ismä'Uiyya (W. Ivanow) in the Encyclopaedia of Islam, Supplemerzt. (To the author of this article, as well as to Hamdani, the pages of Wiet were evidently unknown.)

1 This seems simply to be the author's way of speaking; at least I do not know of any discussion of the historicity, or otherwise, of al-Tayyib, either by Arabic or modern historians. What is meant is, obviously: whose existence is open to reasonable doubts. - The same reservations about the existence of al-Tayyib are expressed by Ivanow also in A Creed of the Fatimids, p. VI: "the pro- blematic infant son of al-Amir, al-Tayyib"; The Rise of the Fatimids, p. 20: "Al-Tayyib ... whose historical reality is highly questionable and whose existence is a matter of faith". 1 See Cl. Cahen, Quelques Chroniques des demiers Fatimids in: Bulletin de 1'lnstitut Français d'Arch6ologie Orientale, 1937, p. 5.

' Cf. on the Great Hall al-Maqrizi, Khitat, I, 388. 2 Cf. for him Wiet, in: J.A., 1921, II, p. 66, n. 5. a Quoted in the Majmuc al-taybiya of Tahir b. Muhammad (see Ivanow, Guide, p. 53); then in the 'Uyitn al-akhbâr, VII, 3 I ff. * Sec on this passage below, p. 218. � It seems that the name Sayyida Arwa given to the queen by Hamdani (in the article quoted above, p. 195, n. x ) is based solely on the corrupt passage iIl'Umara, Kay p. 22, where the word"Arwa" must be deleted. Arwa was the name of a daughter of the queen (Kay pp. 76, x6o, 2g6). The queen is always named al-Sayyida, which is expressly stated to have been her personal name, `Umara, Kay p. 28: ismuha <al-> Sayyida bint Ahmad etc.; 1[7yfin al-akhbir, VII, 134: wa-zawwajahu (scil. al-Mukarram) abuhu al-lturra al-taqiyya al-Sayyida ibnat Ahmad etc.; al-Malik al-Ashraf Ibn Rasul: Turaf al-ashdb fi malri/at al-ansdb, ed. Zettersteen, Damascus 1949, p. rx7: al-Sayyida bint Ahmad, ma-hiya al-Sayyida bint Ahmad b. Muhammad b. al-easim b. Mu:;atfar, wa-qäla nås: b. Muhammad au b. calm wa-AUdh a`lam. - About al-Sayyida as a proper name cf. C. A. Nallino, Raccvlta di Scritti editi e inediti, vol. III, p. 433.

1 The passage from the lost chronicle of Ibn Abi Tayyi has been preserved by Ibn al-Furat;

1 These persons do not seem to be known from other sources; the pronunciation of the names of two of them is uncertain.

1 It is certain that the role attributed by Idris to al-Hasan the son of al-Hafiz is due to a con fusion with the events of the year 529 (see the chronicles under that year).

1 Quoted by Ibn Taghribirdi, ed. Popper, III, q-g. Cahen, p. zo-m, has shown that Ibn Khal- dun's account, IV, 7o ff.� is a condensation of that of Ibn al-Tuwayr. - For references to the other historians dealing with these events cf. Wiet's pages quoted above, p. 195.

1 It is obvious from the account of Ibn al-Tuwayr and the other chronicles that the notion of the 'Uyun al-akhbdr about 'Abd al-Majid initially assuming the regency on behalf of al-Tayyib is completely false. The Tayyibi tradition puts the blame for the ousting of al-Tayyib squarely on Abu 'All Kutayfat. As a matter of fact, the claims of al-Tayyib were discounted from the very be- ginning of the regency of 'Abd al-Majid.

1 According to Ibn al-Muyassar (p. q4) on Monday (or Thursday) the 16th of Dhu-1-qa'da. "He arrested Abu-1-Maymun ('Abd al-Majid) the morning after his bay`a". The last phrase is not quite clear; does it mean: the day after he, scil. the new vizier, had payed his allegiance to the regent, 'Abd al-Majid? or does bay'a here mean simply: his appointment as vizier? ' This seems to be confirmed by the fact that the khutba was read in the Yemen, as we shall see, in the name of al-Tayyib (this being due to the special position of the Yemenites), cAbd al- Majid as regent and Abu 'Ali as vizier.

1 According to al-Ruhl "he dropped the mention of the House of Ismacil" (i.e. the Fatimids), asqat dhikr dl Ismä'il. The same information occurs also in Ibn al-Muyassar (p. 75) in an altered form: "He dropped the mention of Ismalil b. jalfar from whom the Ismdliliyya derive their name". The person of Isma'il does not play any special role in Ismaili ritual - so this form of the notice is obviously corrupted from that which it has in al-Ruhi. - According to al-Ruhi he abolished from the adhkn the formula Muhammad, re·a-`Ali khayq al-bashar; Ibn al-Muyassar mentions, instead, the formula hayy '■aid khayr at-'amal. * It is probable that the persecution of the staunchest adherents of al-Tayyib, mentioned in the passages of Ibrahim al-Hamidi, has nothing to do with the adoption of Imamism; it probably belongs to the preceding period and was due to their opposition to the regency of 'Abd al-Majid on behalf of the unborn child of al-Amir, instead of al-Tayyib.

1 The following account of the accession of al-Hafiz by Ibn Hammad (548-628) - which he received orally from the qadi Abu-1-Makarim Hibat-Alldh al-Misri - though it is rather confused, deserves to be quoted (ed. Vonderheyden, p. 60). "Al-Hafiz was of the family of the Caliphs, but was not the son of a Caliph, while it was part of their law and doctrine (read: wa-kdn min rasmihim �a-vnadhhabihim all6i yltwallü ... instead of wa-kdna-bna ra'isilzim wa-madhhabuhum ... ?) not to entrust the Caliphate to any but the son of a Caliph. Al-Amir died childless, but it was thought that he left one of his wives pregnant". The next words are, no doubt, corrupt: wa-lidhalika bi-l- lJàfi;. "As no child was born, �Abd al-Majid al-Hafiz said to his ddci - this is how they name the

. This is confirmed by the declarations made in the anti-Nizari demonstration of 516 A. H. and the corresponding passage of the al-Hiddya al-flmiriyya. The argument was there used to prove the superiority of al->\4usta`li, that he bore the title of wali '�M al-mu'minin, while his brothers (among whom is Abu-1-Qasim, the father of al-Hafiz) had to content themselves with the inferior status of wali `ahd al-muslimin (see the article quoted above p. 194, n. i, p. z3).

In the inscription, instead of al-qddi al-fsau'ayyad al-amir Sirdj al-din, as Wiet, read: al-qddi al-mu'ayyad al-amin S. 2 AL-daula al-hdfiziyya and al-imdma al-fydfiziyya occurs also in documents emanating from the Chancery (cf. al-Qalqashandi, VI, 450, VI, 108; VIII, 344 - quoted by Wiet, who seems, however, to read too much into this usage). 3 Siraj al-din Abu-1-Thurayya Najm b. jalfar, qadi al-qacdkt and dd'i al-ducdt, appointed in 526 (Ibn al-Muyassar, 76' 9); Abu-1-Zahir Isma'il b. Salama, in office in 543 (idem, 88', cf. 75"'); Abu-1- Zahr b. lsmacil b. cAbd al-Ghaffar, in office in 549 (idem, 931'-'a); al-jalis Ibn 'Abd al-Qawi, before and during the times of Shirkuh (al-Maqrizi, Khita,t, 1,391; according to this passage, during the last times of the Fatimid dynasty the Chief Da'Iship was hereditary among the Banal cAbd al- Qawi).

1 About the pretender in Fes (executed in 600) and his son (610 or 612) cf. also R. l3runschvig, in: Melanges Gaudefroy-Demombynes, p. 156, note i. 2 More correctly: in which the poet 'Unizira, too, was implicated.

1 Published by H. C. Kay, London 1892. In the only manuscript known the book bears the title Tca'riklt al-Ya�nart. So also in Abu-1-Fidal, Cairo 1325, I, 3 (Introduction); II, 24 (s.a. 203); II, 153 (s.a. 412); III, 35 (s.a. 554). In Ibn Khallikan, s.v. 'Umdra (ed. Wüstenfeld V, 104), and s.v. 'Ari b. Muhammad al-Sulayhi ( V, 91) it is called Akhbdr al-Yama9a. Other sources call it Kittib al-Mufid fi akkbar Zabid: al-Janadi, in: T>ereubourg, 'Ozimara du Yemen, II, 547; op. cit. 544 has: D�ufid 'Umara, in order to distinguish it from another book of the same title, the If. al-W. fi a. Z. of Jayyash; op. cit. 542, f>3r, 633, 637, 649: al-,,Ifttlid; Al-Sakhawi, 7%[M al-tanbih, p. 127: al- lllu/ad fi akhbar Zabid; Idris, �!7yMtt al-akhbar, passim: �-Mtt/t�. The history of Zabid does, in fact, occupy a prominent place in the History of 'Umara; nevertheless, the second part of the title: "... Ii akhbar Zabid" does not seem to be altogether justified. 2 Cf. I. Goldziher, Streitschrift des damli gegen die Bdtinijja-Sekte, p. 7-8.

1 Brit. Mus., Or. 3265 (Rieu, Supplement, p. 374-8). The manuscript is very faulty; nor is Kay's treatment of the text satisfactory. (His merit lies in the rich material brought together by him for the historical commentary.) Some emendations have been proposed by W. Robertson-Smith in J.R.A.S., 1893, p. 18 ff.; cf. Kay's rejoinder, p. 218 ff. Since the time of Kay's edition, much material, in the form of quotations from the book, has become available, so that a much better text could now be produced. - As an illustration of the state of the text, I wish to dwell at some length on one particular passage that has already exercised the minds of several scholars. On p. 58 'Umara tells a story about the generosity of the dd`i `Imran b. Muhammad, the Zurayi'id. He (`UmSra) owed the prince the amount of three thousand dinars; but on returning to his court from Egypt, gave him a letter from the Fatimid vizier al-b4alik al-Salih (Talä'i' b. Ruzzik) de- manding that 'UmAra should be allowed to pay back the money by instalments. (That is the meaning of the term taqsit, cf. Dozy, II, 344; it is misunderstood by the translator.) "What is the purport of al-Malik al-Salilo's letter" asked the dii'E. Al-Qadi al-Rashid, who was also present, answered: taqsit. The ddH answered: k_¥ cJ'L¡\\ je ia— 1 � fJi W Kay 'emends' (note 2): 1�'S 4j\�\ Je � and translates: "Let 'Umara offer us two lines of verse, in which he shall strictly observe the laws of rhyme, and the account shall be regulated." W. Robertson Smith, in: J.R.A.S., 1893, p. 206-7, after a suggestion about the meaning of taqsit (which is, however, wrong), remarks: "The first 1:.--:. seems to be a dittography. "Let him first produce two verses on the subject rhyming in qdi." The prince prescribes the rhyme to secure an impromptu." H. Derenbourg (Oumara du Yemen, II, in) is not much nearer to the point: "Let ''Umara recite, exclaimed the dä'ï, two verses ending in the letter qdf preceded by sin". In a note, Derenbourg adds: "The command of the dä'i, about the rhyme in sin qäj, results from the form in which the passage of `Umara is quoted in Ba Makhrama, Ta'rikh, fol. 93vo. [The text as it is quoted by Ba Malchrama - whom Derenbourg, adduces to support his interpretation - bears out mine; see cd. Lofgren, vol. II (ig5o), p. 184.] Sin precedes qdf in the verb saqata, 'to fall', which I suppose must be read wherever Kay printed qasata ... cf. Kay, Yamarv, p. 58 note 2. In the translation, p. 78, certain details are open to doubt, but the passage as a whole is clear." - The correct interpretation of the passage is, no doubt, as follows. Read: lM.—jL' �. �� r-� L (� and Oy- represent, of course, the same skeleton: 0'-)' Translate: "Let the sin of the word ja— ("it shall be paid by instalments") be made to precede the qdf of the same word, so that it (becomes) izi_j ("it - scil. the debt - shall be waived")".

1 Ivanow gave further details about the treatise in: Journal of Bombay Branch of R.A.S., 1939, p. i ff. I have no manuscript of the book at my disposal; the passage in question is reproduced from the extract in the Kitab al-azhvar. See Appendix no. II. The History of ''Umara, quoted under the name of al-Mufld (cf. above note 52) is one of the main sources of vol. VII.

1 Sulayman b. 'Amir, of the eminent Ismaili family of al-Zawahi, half-brother of al-Malika al- Sayyida; his appointment as dia`i is mentioned also in the body of `UmSra's book, where it is added that he was subsequently poisoned by the Queen's minister, al-Mufaddal (p. 28; cf. also P. 34). 2 There is a lacuna in the manuscript between the words qddi and b. Malik; there is, however, not the slightest doubt that we have to supply: Yahyd b. Lamas. On him see Ivanow, Guide, p. 50, and below in the present study, p. 219. ' His history is told at length in the body of 'Umiira's book (p. 42-8). All the information of the 'Uyûn al-aklabdr about him (VII, 297 ff.) derives from `LJmara. (ldris quotes, however, in addition two poems by Muhammad b. Ahmad b. 'Imran praising the exploits of Ibn Najib al-daula; p. 304-5.) - I wish to put forward, with due reservation, the suggestion that this 'All b. Ibrahim b. Najib al-daula was a grandson of Najib al-daula 'All b. Abmad al-Jarjara'i (mentioned in Ibn Taghribirdi, ed. Popper, II, 130, 142). - On the ultimate fate of Ibn Najib al-daula, about which 'Umdra and the other Yemeni historians have no information, cf. Ibn al-Muyassar, p. 20 (1. m read: ..1.,...\\ �� 2r; Ibn Khaldun, IV, 70; F. Wiistenfeld, Patimiden, p. Z96ff. (after Ibn Zafir). "'Urndra's silence on the subject is easily understood. Tielieving the dä'i to be innocent, he did not wish to charge al-Amir with a judicial murder" (Robertson-Smith, loc. cit., p. 205).

1 See Kay, note ro2, p. 297-300. - Kay has seen some of the difficulties of the text of 'Umdra. He points out, very pertinently, that "the unsatisfactory manner in which the edict of the khalifa al-Amir is introduced and interrupts the subject of the chapter, is in itself suspicious". Kay will not be blamed if some of the further suggcstions in the note miss their mark; the background of these events was, of course, quite obscure at the time.

1 As the British Museum manuscript of cUmara can have hardly belonged to Ismailis, we must, the hypothesis of an interpolation admitted, assume that the archetype from which it was copied passed through Ismaili hands. One takes recourse to the assumption of so far reaching an inter- polation with diffidence - but I cannot see how one can escape the force of both the internal and external evidence.

1 Note the discrepancy; while in the first paragraph it is implied that al-Sayyida did not exercise a spiritual function, here she is called hujja. It is suggested that the difference is between the genuine 'Umdra, who does not know of al-Sayyida as a spiritual authority, and the Tayyibi tra- dition which does (confer below p. 221, 223, 224, n. 3). 2 The text of the sijill can be corrected with the help of the version contained in the `Uyun al-akhbiir (cf. p. r97, n. 3). P. 100 1. 17 read klialisat al-Imdm instead of klsrzssat al-Imdm. L. T8 al-mustajibin is, of course, correct, al-mustanjidin a mistaken alteration. L. 21 after yalzmadu add i7ayka. P. 101 1. 2 muhtadin, '!7yMK: aL-rnahdiyyiTC, probably correctly; I. 4 after inqadk add minhd; 1. 5 a'-qabahd read -hf4; 1. 8 after zakiyyan add radiyyan; 1. 11 'MXsM&t read ghurratihi; 1. 12 al- mafasil read al-fado'il; 1. 14 after qadaiaa add bihi; 1. 21 �!7yM)t reads, more correctly: wa-ghauthan li-l-mustarikhin wa-ghaytlaan li-l-muntaji`in instead of wa-`auraan lt-L-m.ulLtayri-n wa-ghauthan Li-l-muntaji`an. P. 102 1. z after mahalliki add 'Mt�aAM; imtanaca, 'Uyfl�t reads irtafa'a; 1. 6 minha read minhum; 1. 7 minhu wa-l-hudur read fi-l-bddin ��a-l-huduy; wa-'lamï read wa-'mali. There are small divergences in the final formulas. 3 This protocol of the Queen recurs also in the khutba quoted below and in the documents pre- served in the collection of al-Mustansir's sijilidt (cf. B.S.O.S., 1933, p. 307 ff.) and in the 'Uyfift al-akhbdr. The information given in this passage about Ibrahim is one of the main difficulties in the whole problem. According to Tayyibis sources (cf. below and Ivanow, Guide, p. 52) he was ap-

pointed as an assistant of al-Dhu'ayb after the death of al-Iihattab (533), succeeded to al-Dhulayb on the latter's death (546), and died in 557. According to our text he was appointed de7ci in 526 and, according to al-Janadl's version (it is difficult to decide what the original text of 'Um5rt was) died before the transfer of the fate's to the Zuraylids, i.e. soon after 526. We have to assume that the Tayyibi tradition is substantially correct. - Cf. below, p. 228.

1 The authenticity of the text is not above suspicioa- cf. the following notes. The reasons for doubt are, however, not decisive and a final judgement will become possible only if and when sources as yet unused yield more materials on which to base our criticism. z This is a much less elaborate protocol than that adopted by Abu 'All Kutayfat later, after his coup d'état, and tallies well with the practice of the preceding viziers. 3 if this title of hujja is more than a formality and implies that al-Malika al-Sayyida stands at the head of the da`wrz, there arises the difficulty that, as we have seen, 'Umara's list proves that

the Queen was not holding the office of da`i. A conjecture that at some time the objections to her sex were overcome and the Queen was appointed as dd'i is perhaps not impossible, although im probable (after Ibn Najib al-daula's fall, there followed as dâ'i cali b. `Abdallah - and this leaves little space for al-Sayyida as da'i); at least there seems to be no evidence to confirm it. All these points (cf. also above, p. 221, 223) can only be clearedup after an examination of al-Khattab's Ghdyat al-mamdlid. 1 One is tempted to doubt the authenticity of the khutba also on the score of the mention of al-Tayyib. On the other hand, the mention of Abu 'Ali may speak for its genuineness. The ex- planation of the situation as given in the text will perhaps be found acceptable.

1 According to it al-Mahdi, the first Fatimid Caliph, was not a lineal descendant of Muhammad b. Ismd'il; cf. B. Lewis, Origins of Isinti'ilism, p. 109; W. Ivanow, Rise of the Fatimid, p. 20-23.

1 The enumeration of the jewellery, although not very rich in technical terms, has some ar- cheological interest.

1 The main source for the history of the Zuray'ids is `L;mara, p. 50-59. The text is very faulty, but can be corrected from Ibn Mujawir, ed. Lofgren, Texte zur Geschichte der Stadt Aden im Tflittel- alter, I, 40 ff., where the whole chapter of 'Umara is cited in a more correct form. (Encyclopaedia of Islam, s.v. Karam (Banu '1-), by R. Strothmann, merely resumes 'Umara.)

The second half of the inscription is completely misunderstood by the editors. It means that 'fmr�in died before he could fulfil his wish of making the pilgrimage - so it was his coffin that was taken to Mecca and buried there. (The last meaning must be concealed by the words _,W\, ■t.i- 4��� , which the editors seem to misread.) That the body of 'Imran was transferred to Mecca is also mentioned by al-Janadi, who quotes a passages to that effect from some book of cUmara (cf. Kay, note 68, where the text is given, however, in a corrupt form; a more correct text is to be found in Ba Makhrama, ed. Lofgren, 11, 186-7). One of the authorities of Ba Makhrama (loc. cit.) - Al-Janadij `iJmara - gives the date 560; another - the book at-Sharaf al-a`la (Brockel- mann, II, 173) - the correct s6i.

1 About the resistance offered to the armies of Saladdin by the last Zuraycids and Hamdanis cf. Kay, notes 63 (p. 274 ff.) and 101 (p. 294 ff.).

1 From here the text is also reproduced in the 'u yÜn al-ak3abar, VII, 166 ff. I This is the word used by Hatim for the venia docendi of a ddli,

1 I have no further information about this person. 2 Here ends the quotation in the `Uyun al-akhbar.

1 Obscure. 2 Cf. Guide, p. 56.

1 Cf. Guide, p. 59. He later became the chief of the dahlia and died in 626. 2 Repeated by mistake ? Or is some part of the name to be changed?

1 Perhaps a member of the great Ismaili family of al-Zawahi, who played a very prominent part in the history of the sect in Yemen. — The great interest of the passage is evident; I shall comment upon it in another connexion. [The MS has Vjk->vJ 4\ �,"3 -� — ..\ being a correction above the line.] 2 Cf. Guide, p. 51.

1 For Jurayb in the country of Hajur see al-Hamdani, Jazirat al-`Arab, p. 113. 2 In the text to be quoted presently: Harith (?) b. Sharahil. The Jurayb quoted by Kay, note 76, has hardly anything to do with the family we are dealing with. According to the same note, al-Khazraji reads in our passage: al-Harith. 3 MS ,)\t-*- „) (�. For the geneology from Hamdan to 'Ar-lb cf. F. Wiistenfeld, Genea- logische Tabellen, Table no. 9. (In that table, however, Hajur is given as H. b. Aslam b. 'Alydn b. Zayd b. Jusham.) Ibn Hazm has (Jamhayat ansdb al-`Ayab, ed. Levi-Provencal, p. 369-70): I3. b. Asl. b. 'Aly. b. Zayd b. 'Arib b. Hashid b. Jusham. Instead of Nauf b. Hamdan (as in our text and Wiistenfeld) Ibn Hazm reads Naufal b. H. Ibn Rasul, Tuyaf al-ashdb, p. 117, gives, in the genea- logy of the Sulayhids: H. b. Asl. b. 'Aly. b. Zayd b. 'Arib b. Jusham al-ausat b. Hashid b. Jusham al-akbar b. Khayran (sic) b. Nauf b. Hamdan. (The last four links also p. 123.) . This piece of information provides an important link in the history of the Zaydite Sulaymdni Imams, which can now be reconstructed as follows. (Kay, note 88, has, as usual, some good sug-

1 See for them Kay, Index. (It is only Abu Muhammad Masrur who does not seem to be mentioned by 'Umara; we have probably to read: Abu Mug. Surur.)

1 He composed a dirge on the death of al-A'azz, son of 'All al-Sulayhi, in 458 A.H. (c[7yfln al-akkbky, VII, 127).

1 It is evident, however, from the verses quoted subsequently that he remained loyal to the Fatimids.

1 It is true that he does not know - or does not mention - that there has been a change of dynasty.