A revised chronology of the Sultans of Kilwa inthe 18th and 19th centuries

http://www.aluka.org/action/showMetadata?doi=10.5555/AL.CH.DOCUMENT.sip200022

Use of the Aluka digital library is subject to Aluka’s Terms and Conditions, available athttp://www.aluka.org/page/about/termsConditions.jsp. By using Aluka, you agree that you have read andwill abide by the Terms and Conditions. Among other things, the Terms and Conditions provide that thecontent in the Aluka digital library is only for personal, non-commercial use by authorized users of Aluka inconnection with research, scholarship, and education.

The content in the Aluka digital library is subject to copyright, with the exception of certain governmentalworks and very old materials that may be in the public domain under applicable law. Permission must besought from Aluka and/or the applicable copyright holder in connection with any duplication or distributionof these materials where required by applicable law.

Aluka is a not-for-profit initiative dedicated to creating and preserving a digital archive of materials aboutand from the developing world. For more information about Aluka, please see http://www.aluka.org

A revised chronology of the Sultans of Kilwa in the 18th and 19thcenturies

Author/Creator Alpers, Edward A.

Date 1967

Resource type Articles

Language English

Subject

Coverage (spatial) Northern Swahili Coast, Tanzania, United Republic of, KilwaKisiwani

Source Smithsonian Institution Libraries, DT365 .A992

Relation Azania: Journal of the British Insitute of History andArchaeology in East Africa, Vol. 2 (1967): 146-163

Rights By kind permission of Azania (British Institute in EasternAfrica).

Format extent(length/size)

21 pages

http://www.aluka.org/action/showMetadata?doi=10.5555/AL.CH.DOCUMENT.sip200022

http://www.aluka.org

A Revised Chronology of the Sultans of Kilwa

A Revised Chronology of the Sultans of Kilwain the Eighteenth and Nineteenth CenturiesbyEdward AlpersDr. Alpers is a lecturer in the History Department of the University College, Dares Salaam. He has been studying the history of the Yao and of the MalawiEmpire; to this examination of the Sultans of Kilwa in the eighteenth andnineteenth centuries he bringsan extensive knowledge of the Portuguese sources.Recent publications by Neville Chittick (1965, 1966), based on hisexcavations atKilwa and on his interpretation of the Portuguese and Arabic versions of thesixteenth century Kilwa Chronicle, as well as on the numismatical evidence,challenge G.S.P. FreemanGrenville's earlier (1962a) reconstruction of the pre-Portuguese history of Kilwa and of the chronology of its sultans. Chittick'scriticisms have generally met with wide approval. Less attention hasbeen paid,however, to Freeman-Grenville's tentative reconstruction of theeighteenth andnineteenth century sultans of Kilwa, which appears as an introductorychapter,entitled "The Sultans of Kilwa, c. 1700 to c. 1856", in his The French at KilwaIsland (1965a, pp. 28-38). In a review of this volume (1965), I briefly pointed outcertain weaknesses in this later chronology; but Freeman-Grenvilleis correct instating (1966, p. 7) that his dating of the rulers of Kilwa in this period "has as yetnot been challenged." In this article I wish to analyse the way in which Freeman-Grenville utilizes the sources at his disposal; then to introduce some new evidencewhich was unknown to him; and finally to propose a revised chronology andgenealogy for these more recent sultans of Kilwa.Still the earliest known relevant document is a letter which was sent toPortugal in1723 by the Sultan of Kilwa, Sultan Ibrahim b. Sultan Yusufb. SultanMuhammadb. Sultan Alawi. According tojustus Strandes (p. 99, n. 15), who firstpublishedthis reference in 1899, this letter was located among the "MS. Liss. Archivo doConselho Ultramarino. Papeis de Servigo No. d'Ordem 1046." Since Strandesundertook his research at the end of the last century in Lisbon, there have beenseveral major relocations and reclassifications of, as well as substantial additionsto, the body of archives then extant there. The business of locatingdocumentsmentioned in early authorities is therefore no simple matter. I have,for example,looked in vain for this document under all the obvious headings in thevariousarchives in Lisbon. Consequently, we still do not know either the contents of theletter or exactly to whom it was addressed. Nevertheless, it makes a usefulbeginning for this sort of reconstruction, especially as the genealogical details areall verifiable from other sources. The next source which was known to Freeman-Grenville is the treaty which was concluded between the Sultan of Kilwaand theFrench slave

140 Sultans of Kilwatrader Jean-Vincent Morice' on 14 September, 1776. Freeman-Grenville initiallycites the information on the sultan's seal to this treaty, which renders only thenames of the ruler and his father, Sultan Hasan b. Sultan Ibrahim, the name of thegrandfather being omitted on the seal. But, he points out, this name is supplied inthe text of the treaty itself, so that we have the more complete name of SultanHasan b. Sultan Ibrahim b. Sultan Yusuf (Freeman-Grenville, 1965a, p. 29).If we turn to the full text of the treaty, however, we find that a much moreextensive genealogy is given in the treaty than that which Freeman-Grenville citesin his introductory chapter. According to Freeman-Grenville's own translation(ibid., p. 71), the treaty opens with the following proclamation: "From the slavewho trusts in God, the Sultan of Kilwa [of the line] of the Kings, Sultan Hasanson of Sultan Ibrahim son of Sultan Yusuf son of Sultan Muhammad son ofSultan Ali the Kilwa Shirazi." Later on in his reconstruction, Freeman-Grenvilleintroduces the Habari za Zamani za Kilwa Kisiwani. (For an Englishtranslation,see Freeman-Grenville, 1962b, pp. 220-226.) From this history we see that duringthe reign of Sultan Hasan b. Ibrahim "a Frenchman came to do business"(Freeman-Grenville, 1962b, p. 223). As Freeman-Grenville notes (1965a, p. 35),it is clear that this is the same ruler who signed the treaty with Morice in 1776.But for Freeman-Grenville certain problems are raised by the factthat this sultan'sgenealogy, beyond his father, is at odds with Strandes' evidence and the(incomplete) information which he himself has gathered from the treaty.According to the Kisiwani tradition, Sultan Hasan's full genealogy isSultanHasan b. Sultan Ibrahim b. Sultan Muhammad b. Sultan 'Ali. Comparing thisruler's descent with the other data, it seems perfectly reasonableto assume that theoral tradition has merely forgotten the name of Sultan Yusuf b. SultanMuhammad, who is mentioned in the letter found by Strandes and in the body ofthe treaty with Morice. Indeed, Freeman-Grenville reaches this same conclusion(ibid., p. 35). But because he ignores Sultan Hasan b. Sultan Ibrahim's fullgenealogy as given in the treaty, Freeman-Grenville forces himself into a positionwhere he must try to account for the inclusion of a Sultan Muhammad in theHabari za Zamani za Kilwa Kisiwani and his omission in the treaty of1776.Although he raises several interesting points in this discourse, all hisefforts to thispoint are unnecessary, as the treaty does include Sultan Muhammad's name.Furthermore, the treaty quite plainly states that Sultan Muhammad's father wasSultan 'Ali. Thus the identification of Alawi with 'Ali seems evident.HadFreeman-Grenville made full use of the sources at hand he should have found noproblems in collating Strandes' evidence with that in the text of the treatyconcluded between the Sultan of Kilwa and M. Morice in 1776. Put in tabularform we can see this quite clearly (see Table opposite).In connection with 'Ali, or Alawi, the earliest documented Sultanof Kilwa in thisperiod, it is worth noting that the Kisiwani tradition refers to a Shirazinamed Alib. Sulaiman as the founder of Kilwa. As Freeman-Grenville points out, there is astrong case for believing that this is merely a late nineteenth centuryrendition ofthe founding of Kilwa as documented in both the Arabic and Portuguese variants

of the Kilwa Chronicle. On the other hand, if the 'Ali b. Sulaiman of tradition is tobe identified as the 'Ali/ 1. Archives Nationales de France, Paris, Colonies s6rieC4, vol. 49: Morice's Christian name appears in the copy of a contract, dated PortLouis, 30 September 1779, for the purchase of six hundred slaves by theGovernment of Ile de France. I am grateful to Dr. Freeman-Grenville for allowingme to utilize his draft paper, "Some Eighteenth-century DocumentsconcerningEastern Africa in the Archives de France." (1965b).

Edward AlpersStrandes TreatyAlawi 'AliI IMuhammad b. Alawi Muhammad b. 'AliI IYusuf b. Muhammad Yusuf b. MuhammadIbrahim b. Yusuf Ibrahim b. Yusuf(F1. 1723)]Hasan b. Ibrahim(Fl. 1776)Alawi of record, I think it is preferable not to pursue this point Cf.ibid., pp.34-35). Rather, I feel that we should accept the fact that telescoping hastaken place,and then concentrate on trying to place the modern 'Ali/Alawi b. Sulaimanin hishistorical context. This is not so easily done. In the Habari za Zamani za KilwaKisiwani, 'Mi b. Sulaiman is said to have married the daughter of Mrimba, thefirst headman of Kilwa Kisiwani. But 'Ali seems clearly to be datable to theseventeenth century, while the only other evidence we possess for Mrimba seemsto point to a mid-eighteenth century dating (Abdallah, p. 9, 11. 9-17). Thisassignation, however, is contingent upon our acceptance of the identification ofMrimba as a unique personality. At present, there is no way to substantiate such abelief, and I am inclined to regard the name Mrimba as an inheritabletide, ratherthan a personal name. This, then, is a point which requires further research.Returning to Hasan b. Ibrahim's date of accession, Freeman-Grenville (1965a, p.81) bases his dating on Morice's testimony, written from Kilwa on 4 November,1776, that:* . . the king who now reigns in Kilwa is the youngest of three brothers. Theeldest of these reigned for two years and was dethroned, ... because he drank toomuch. The second only reigned for a year; he was deposed because he was of toolimited an intelligence. The third has reigned for four years, everyone is delightedwith him. It seems that he will go on reigning until the end of his life. These twodeposed kings are both on the island andhave no title but that of private citizens.On the unverified strength of this evidence alone, Freeman-Grenville (ibid., p. 30)dates the death of Sultan Ibrahim b. Yusuf and the accession of his eldest son tothe throne in 1769. Similarly, he ascribes this son's deposition fordrunkenness to1771; and a year later he has Sultan Hasan b. Ibrahim b. Yusuf taking over fromhis other unqualified elder brother.

Were Morice our only source on this matter, we would have little alternative toaccepting what he says. In fact, the Habari za Zamani za Kilwa Kisiwanimentions a younger brother of Sultan Hasan b. Ibrahim, who preceded him assultan. But the details of his reign seem less reliable than Morice's information,for we are told that when he came to power "there was trouble in the land forseven years from famine and locusts*... We know nothing of the events of SultanIsufu (Yusuf), because he did not reign properly on account ofthe trouble"(Freeman-Grenville, 1962b, p. 223). The suspicions that are

148 Sultans of Kilwaraised by the assignation of a seven year reign to this ruler are confirmed by asimilar passage in the Swahili History of Pate, where we learn that a certainSultan Ahmad "reigned seven years without rain falling, and then he abdicated ofhis own free will..." (ibid., p. 258). In these circumstances, Freeman-Grenville isquite justified in prefering Morice to the Habari za Zanani za Kilwa Kisiwani.Regrettably, in his tentative reconstruction, he completely ignores Strandes'reference (p. 302) to the fact that in 1759 Sultan Hasan b. Sultan Ibrahim b. SultanYusuf sent an envoy to Mogambique with news of the struggle betweenMombasa, Pate, and Muscat, and assuring the Portuguese of his continuedfriendship. The omission of this valuable piece of information becomes all themore remarkable when we find Freeman-Grenville citing this very same referenceon page 41 (1965a), just three pages after the presentation of his final tentativereconstruction. Furthermore, Strandes' evidence also indicatesthat Morice cannotbe regarded as an infallible source of information. We must therefore rejectFreeman-Grenville's dates for the death of Sultan Ibrahim b. Yusufb.Muhammad, for the reigns of his two elder sons, and for the accession of SultanHasan b. Ibrahim b. Yusuf.New archival sources have recently been found in Lisbon which sustainthisjudgment; but before turning to them it is necessary to continue ourexamination of FreemanGrenville's reconstruction in The Frenchat Kilwa Island.Moving towards the early nineteenth century rulers of Kilwa, I am unable to addanything to Freeman-Grenville's comments (ibid., pp. 30, 36) on theuniquereference (Gray, 19-62a) to Sultan Abu Bakr b. Sultan Hasan b. Sultan Ibrahimthe Shirazi of Kilwa, who wrote a letter to the Governor of Ile de France in 1797.This sultan's existence, his dating, and his pedigree are indisputable. Yet he isnowhere mentioned in the Habari za Zamani za Kilwa Kisiwani.FreemanGrenville's placement of this individual as Sultan Hasan b. Ibrahim'simmediate successor (if not necessarily his eldest living son) is the only way inwhich we can make sense of this reference.With the last four sultans of the line Freeman-Grenville has less success. Here,again, he fails to utilize the available sources to their fullest advantage, whiledrawing some rather dubious conclusions from the same sources. The first ofthese rulers, Sultan Yusuf b. Sultan Hasan b. Sultan Ibrahim, is said, in the Habariza Zatnani za Kilwa Kisiwani, to have been his father's eldest son and successor.Perhaps this is an indication that the reign of Sultan Abu Bakr b. Hasan b. Ibrahimwas not a happy one, and that his elder brother, Yusuf, soon replaced him as

sultan. In any case, whenever Sultan Yusuf b. Hasan came into office, it isvirtually certain that he was reigning when H. M. Frigate Nisus visited Kilwa in1812. We owe this information to the diligence and curiosity of the ship'ssurgeon, James Prior, who records that the sultan's name was Yousou Fou (p. 68).Yet Freeman-Grenville persists in dating this voyage, and thus the onedate wehave had up to now for Sultan Yusufb. Hasan, to 1811 (1962b, p. 202; 1965, p.28, 36), although Prior himself states quite explicitly, in the Advertisement to hisjournal (p. iii), that "the voyage along part of the Eastern coast ofAfrica... tookplace in consequence of the arrival at the Cape of Good Hope, in 1812, of anembassador (sic) from the King of the Comoro Islands to the governor of thatcolony.. ." This point also seems clear enough to Sir John Gray (1963, p. 222).Sultan Yusuf b. Hasan b. Ibrahim, or at least Prior's Sultan Yousou Fou, was onthe throne of his ancestors in 1812.For the remainder of his reconstruction, Freeman-Grenville reliesexclusively onthe Habari za Zamani za Kilwa Kisiwani, interpreting its details in light ofhistorical

Edward Alpersknowledge of the coast during this period. He is probably right, I think,inrejecting the possibility (1965a, pp. 36-37) that Sultan Yusuf b. Hasan's son,Mfumo Hasan (Formo Sani), who greeted Prior on his visit to Kilwa in1812 (p.70), ever succeeded to the throne. All the more reliable available evidence pointsto this conclusion (see below, p. 152). But in trying to date Sultan Yusuf b.Hasan's death and the regnal dates of his successors, he encounters a considerableamount of difficulty. In regard to the first point, Freeman-Grenville seizes uponthe statement in the Habari za Zamani za Kilwa Kisiwani (Freeman-Grenville,1962b, pp. 223-224) that "there was great friendship" between Sultan Yusuf b.Hasan and Sayyid Sa'id of Zanzibar, and that "Said bin Sultan was on excellentterms with the people and the ruler." There is nothing unusual in this passage,which seems a common enough accounting of the fact that, as Freeman-Grenvillehimself puts it (1965, p. 32), there were "cordial relations" between Sayyid Sa'idand the vassal Sultan of Kilwa, Yusufb. Hasan. But Freeman-Grenville reads farmore into this passage. "Since," he writes, "Sayyid Said's first visit to Zanzibarwas in 1827, the friendship of the kind described could hardly have occurredbefore that date." He then goes on to suggest that "Yusuf bin Hasan's friendshipwith Sayyid Said should presumably be placed c. 1827-32, if not later" (ibid., pp.32, 36). It must be stated categorically that there is not one shred ofevidence inthe Habari za Zamani za Kilwa Kisiwani that these two men ever met personally.It is also a disconcerting matter of detail to find Freeman-Grenville incorrectlydating Sayyid Sa'id's first visit to East Africa to 1827. Both Gray (1962b, pp. 123-124) and Coupland (pp. 273-274) indicate that although the fleet set sail fromMuscat late in 1827, it did not actually reach East Africa, at Mombasa,until thefirst week of the new year, January 1828.There is, on the other hand, a very clear statement on the chronologicalcircumstances of Sultan Yusuf b. Hasan's death in the same Swahili history. ThisFreemanGrenville disregards. The last paragraph in the tradition which deals with

the reign of Yusuf b. Hasan describes in relative detail the raids made on KilwaKisiwani and on the lesser islands in the Mafia group by the Sakalava people ofnorth-western Madagascar. When the Sakalava had finally beendefeated, andtheir captives returned to Mafia, we are told: "Then Sultan Isufu died" (Freeman-Grenville, 1962b, p. 224). Sir John Gray (1963, p. 222) has previously noted thissituation, adding that the Sakalava raid on Mafia "took place in about 1822." Infact, this date is rather too late; but this is of no significance in the presentcontext.What matters is that Freeman-Grenville could have arrived at a much moreplausible date for the end of Sultan Yusuf b. Hasan's reign than he has, simply bymaking better use of the sources at his disposal. Had he placed Yusufb. Hasan'sdeath at sometime shortly after 1822, he might also have avoided his speculationsconcerning the sultan's friendship with Sayyid Sa'id.What, then, of the succeeding sultans? The Habari za Zamani za KilwaKisiwani(Freeman-Grenville, 1962b, p. 224) has this to say:Sultan Mohamed bin Sultan Hasani succeeded. The peopleintrigued with Sultan Selimani bin Sultan Hasani, to such an extentthat they bothwent to Muscat about the succession-from which Sultan Selimani returned. Whenthey arrived at Muscat before Said bin Sultan, he imprisoned SultanMohamed,and SultanSelimani returned to come and rule himself.

150 Sultans of KilwaWhen he arrived at Kisiwani, the townsfolk sent him back toMuscat, saying: 'Since you wish to rule, release your brother Mohamed fromprison, and bring him here: then you shall rule.' Sultan Selimani returned toMuscat and released his brother Sultan Mohamed, and they bothcame back as faras Zanzibar.There Sultan Selimani died, leaving Sultan Mohamed.The details of the succession dispute between Sultan Yusuf b. Hasan's (probablyyounger) brothers, Muhammad and Sulaiman, should be readily apparent. Theirrivalry reached such a state that either the citizens of Kilwa, or the Omanigovernor, who was the effective ruling power, sent them off to Muscat to availthemselves of Sayyid Sa'id's arbitration. Thrice it is stated that Muscat was theirdestination; once that Sayyid Sa'id was in residence there. It was only afterSulaiman's unsuccessful bid to foist himself off on the people ofKilwa as theirlegitimate sultan, and his return to Muscat to retrieve his brother, that they aresaid to have spent any time at Zanzibar (cf. Gray, 1963, p. 222). Nor is it stated,or implied, that Sayyid Sa'id was then at Zanzibar. The death at Zanzibar ofSultan Sulaiman b. Hasan, by fair means or foul, left Sultan Muhammad b. Hasan,albeit temporarily, as the sole claimant to the throne. The most likely conclusionto be drawn from this evidence is that these appeals were made to Sayyid Sa'idsometime after Sultan Yusuf b. Hasan's death in the early 1820s, andbeforeSayyid Sa'id's coming to Zanzibar, which took place between the time ofhis firstvisit in 1828 and the final removal of his capital to East Africa in 1840. Indeed,given the nature of the evidence, it seems most probable that the dispute and thearbitration all happened before Sayyid Sa'id's first voyage to East Africa, although

we cannot conclude this categorically on the sole basis of the Habariza Zamani zaKilwa Kisiwani. It can, in fact, be ascertained from other sources that this was thecase.But Freeman-Grenville (1965a, p. 36) sees things differently:Sayyid Said's permanent residence in Zanzibar began in 1840.In recounting the quarrel which took place about the succession after Yusuf binMasan's death, although it is not explicitly stated, it appears to be assumed thatSayyid Said was in normal residence in Zanzibar: the tradition speaks of bothSultan Muhammad and his rival brother as returning from Zanzibar after theirappeal. The incident, and Sulaiman's almost immediate death, can thus beplaced in the fifth decade of the nineteenth century.It is clear that Freeman-Grenville has substantially misread the tradition, when hededuces that Sayyid Sa'id was in residence in Zanzibar; and his dating of thisincident must therefore be rejected out of hand. It should be noted, too, that SultanSulaiman b. Hasan did not return to Kilwa from Zanzibar with his brotherMuhammad.Finally, Freeman-Grenville's account of the end of the Kilwa sultanate isunsatisfactory. The last known Sultan of Kilwa seems to have beenHasan b.Sulaiman, who assumed his father's mantle after the latter's death at Zanzibar andpressed his claim to the throne. There are two different traditional accountsdescribing how Hasan b. Sulaiman became sultan and how the sultanate endedwith his reign, while a third variant recounts a similar situation. TheHabari zaZamani za Kilwa Kisiwani (Freeman-Grenville, 1962b,

Edward Alperspp. 224-225) reveals the growing importance of the town of Kilwa Kivinje,situated on the mainland some sixteen miles to the north of Kilwa Kisiwani.Itnotes that the people of Kisiwani warned Sultan Muhammad b. Hasan not to passby way of the rival town on his return from Zanzibar, but that he succumbed tothe invitation of the Kivinje people to do so. Consequently, the Kilwa Kisiwanifolk proclaimed Hasan b. Sulaiman, who was undoubtedly waiting inthe wings,as sultan. The mainlanders and the islanders united behind their chosen sultans,each of whom seems clearly to have been no more than a symbol of theincreasingly bitter rivalry between the two towns, Kilwa Kisiwani struggling forsurvival and Kilwa Kivinje aiming to emerge as the most important trading porton the coast of southern Tanzania."So they made war, and the people of Kisiwani came to Mjengera (justnorth ofKilwa Kivinje) and fought with them for many days.Then Sultan Mohamed made his home at Mkondaji (on themainland) and Sultan Hasan made his home at Kisiwani, until Sultan Mohameddied. And since Sultan Mohamed and Sultan Hasan ruled there has been no properkingdom, nor have we heard of their habits and customs. After the death of SultanMohamed power was in the hands of Said bin Sultan himself. Sultan Hasanhad no power but only the honour of the title Sultan."

Thus, the Habari za Zamani za Kilwa Kisiwani would have us believe that the linesimply expired, for no further mention is made of the sultans of Kilwain thechronicle.The second variant comes from P. A. Lienhardt (cited in Freeman-Grenville,1965a, p. 37), who collected it at Pande, a small mainland village near KilwaKivinje, from "the hereditary circumciser of the survivors of Kilwa'srulingfamily." Lienhardt was told that there were two claimants to the throne. One wasthe previously encountered Sultan Hasan b. Sulaiman; the other was a certainIbrahim b. Mawahibu, who is otherwise unknown to us. The Sultan of Zanzibar(presumably Sayyid Sa'id, but he is not mentioned by name) was asked toarbitrate and decided in favour of Hasan b. Sulaiman. The rejectedIbrahim b.Mawahibu, together with his son Fadhila, left the island and settled on themainland, where Fadhila maintained the family opposition to Sultan Hasan b.Sulaiman after the death of his father, Ibrahim. According to the tradition, Fadhilahad one trump card to play, for the unsuccessful arbitration had granted his fatherthe right to use "the horn of Kilwa, the principal emblem of the sultanate and ofthe Shirazis." Fadhila managed to inveigle the royal horn (of which this traditionis the unique source of knowledge) from Sultan Hasan, who only surrendered itreluctantly.Having gained possession of the horn, Fadhila refused toreturn it and he told his followers that he meant to 'kill' the horn and put an end tothe sultanate for ever. And so, as it is said, Fadhila took the horn anddrowned itin the sea, and as his boat was passing Kilwa he sent a message to Hasan the lastof thesultans, saying:The giant rock cod has swallowed the horn. There is no knowing whetherhe will vomit it up again.Utilizing these two accounts, the second of which is clearly symbolical, telling usnothing at all about the actual fate of Sultan Hasan b. Sulaiman, Freeman-Grenville

152 Sultans of Kilvaconcludes that the Habari za Zamani za Kilwa Kisiwani "is correct in describingSultan Hasan as a feeble nonentity." This much is evident. What is lacking,however, is any attempt to go beyond the facade of the two traditions inorder todetermine the real fate of Sultan Hasan.I mentioned before that there are three versions of how the sultanate came to anend. The last is the Habari za Zamani za Kilwa Kivinje (Velten, pp. 253-264), towhich FreemanGrenville gives unduly short shrift in his reconstruction. For whileit is true that the tradition "represents a very diluted and unreliable form of theHabari za Zamani za Kilwa Kisiwani," (Freeman-Grenville, 1965, p. 34) theaccount which it gives of the fate of the sultanate is a valuable supplement tothose given in the other two traditions. Briefly, the Kivinje chroniclesays, quiteinaccurately, that a certain Yusuf b. Hasan established the Shirazi dynasty ofKilwa Kisiwani. On his death there was a succession dispute between his son anddesignated successor, Hasan b. Yusuf, and his jealous brother, Isma'il b. Hasan.

Their rivalry caused a civil war amongst the islanders, from whichIsma'ilemerged victorious. But his reign lasted only a few days, because Hasan poisonedhim. So it was that Sultan Hasan b. Yusuf came to rule Kilwa Kisiwani andKilwaKivinje, which is here mentioned for the first and only time in the tradition. SultanHasan was popular with his subjects and his reign was peaceful. There were greatprofits from the ivory and gum copal trade during his time. Then, we are told,"Sultan Said bin Said (sic) came from Muscat, and arrived at Unguja. And at thattime, when he came to govern Unguja, there were no sultans in these countries ofKilwa, because all the sultans had died" (Velten, pp. 258-259).The parallel between these events and those described in the Habari zaZamani zaKilwa Kisiwani is striking. Although the names of the rivals are different, itseems possible that the circumstances of Ismail b. Hasan's death by poisoningmay reflect on those of Muhammad b. Hasan's otherwise obscure demise. Anothertantalizing point which is raised by the Kivinje tradition is the unique appearanceof a Sultan Hasan b. Yusuf, who, if the name is indeed correct, mightbe Prior'sFumo Hasan. These are intriguing items of speculation, but no more than that.The really important point which emerges from the Habari za Zamani za KilwaKivinje is that we have yet another totally different explanation given for thetermination of the sultanate itself. Comparing the three very different versionsof this extremely important development, it seems to me that we are confrontedwith three independent attempts to hide the real truth of the matter. Sir JohnGray's researches suggest that this was indeed the case. In hisvaluable History ofZanzibar (p. 183) he writes of Sayyid Sa'id's concern at reportedFrench attemptsto establish a foothold on the mainland (which in December 1841 had beenproved by Atkins Hamerton, the British Consul at Zanzibar, to be groundless), inthe following terms:Later in 1842, however, the French brig-of-war Messagervisited Kilwa and approached certain 'chiefs' there regarding the purchase of theirisland. The chiefs had hastened with the news to Zanzibar and were reported byHamerton to be 'terribly frightened.' Seyyid Said was equally alarmed. Fearingthat the Shirazi Sultan of Kilwa might succumb to the temptation of French gold,he caused him to be deported to Muscat.

Edward AlpersHere, then, in a source which in other instances is utilized by Freeman-Grenville,is the ultimate fate, together with its approximate date, of the Kilwa sultanate. (cf.Burton, II, p. 366, who names the deported last Sultan as Muhammad[Muammadi].)If Freeman-Grenville's tentative chronology of these later sultansof Kilwa isunacceptable, what evidence have we, beyond the texts which he hadat hisdisposal, from which to construct a more realistic outline of this sort? Strandes'earliest notice of Sultan Hasan b. Sultan Ibrahim b. Sultan Yusuf is dated atMogambique, 24 July, 1759 (p. 321, n. 10). But this letter merely refers to thesultan as a third party.2 I have, however, discovered the originalPortuguesetranslation of the sultan's letter, as well as the Portuguese reply to this letter. Thefirst was received at Mogambique from Sultan Hasan b. Sultan Ibrahim b. Sultan

Yusuf (Sultan Assane Bonusultan Ibraimo, Bonusultan Issufo) oneither 20 or 24April, 1759. It tells of the previous year's expedition sent from Muscat against thecombined rebels of Mombasa and Pate (for details, see Alpers, 1966, pp. 154155),and assures the Portuguese governor (then Pedro de Saldanha de Albuquerque) ofcontinued amity between Kilwa and the Portuguese. The letter also explains thatits bearer is the Sultan's brother and ambassador. In his reply, Saldanha deAlbuquerque mentions the Sultan's brother by name: Mwinyi Juma MwinyiJamoto (:) (Moenha Juma Moenha Jamote). This is the only known reference tothis brother of Sultan Hasan b. Ibrahim.3 A year later Saldanha de Albuquerquereceived another, unlocated, letter from Sultan Hasan b. Ibrahim, whichapparently contained further news of the northern coast and expressed the Sultan'sdesire for continued trade between Kilwa and Mogambique.4Five years later, in April 1765, another delegation from the same Sultan of Kilwawas received at Mozambique, as we know from the copy of a "Letterfrom theQuiloa Sultane Assane Bune Sultane Ibraymo, and jointly all the Moors ofMindano [Mikindani?], and Mombasa, and Muenha Combo is able to givethenews of what is happening here."5 The contents of this letter, Mwinyi Kombo'spretences to the throne of Mombasa, the ensuing farcical Portuguese expedition torecapture Mombasa, and Mwinyi Kombo's eventual fate, have all been dealt within some detail elsewhere. (See Strandes, pp. 302304; Freeman-Grenville, 1965a,pp. 220-224; Teixeira Botelho, pp. 491-494; Hoppe, pp. 52-59). The principalinterest of the letter for the present purposes is that it identifies Sultan Hasan b.Ibrahim as a leader among the traditional Swahili coast rulers who were chafingunder the yoke of Omani rule, and seeking Portuguese aid in the cause ofindependence from the Arabs. During the southwest monsoons of 1770, anunknown French slave trader at MoQambique reported to J. Brayerdu Barre, ashipowner at Ile de France, that the King of Kilwa, "a little Moorish town" to thesouth of Mombasa, "has assured me at Mozambique, that he desiresgreatly to beput under the protection2. Arquivo Hist6rico Ultramarino, Lisboa, C6dice 1317, fl. 62-63, Pedro deSaldanha de Albuquerque to Secretary of State, Mogambique, 24July1759.3. A.H.U., Morambique, caixa 7, "Verso da Carta do Rey do Quilloa," n.d., andreply, Mogambique, 27 April 1759; copies in A.H.U., C6dice 1317,fl. 72-73.4. Ibid, fl. 215, Saldanha de Albuquerque to King of Kilwa, Mogambique, 12May, 1760.5. A.H.U., Mogambique, caixa 11, Sultan Hasan b. Ibrahim to Governor ofMogambique (copy by Francisco Pereira de Almeida, dated Moqambique, 15August, 1765), enclosed in Baltasar Manuel Pereira do Lago to Crown,Morambique, 19 August, 1765.

Sultans of Kilwaof the French.. ."6 It is most unlikely that the sultan, who passes without name,was actually at Mogambique in person, but we have seen that it was not unusualfor his ambassador to be there. It also seems reasonable to interpret the sultan'sinvitation for French protection as yet another feeler for militaryaid against theOmani Arabs. This tactic would be in character for Sultan Hasan b. Ibrahim. The

next year we have more positive evidence that the sultan was seeking aid fromevery possible quarter.Sometime before 12 April, 1771, another ambassador from the Sultan of Kilwaarrived at Mogambique bearing this extremely interesting letter from hismaster:I, the King of Kilwa, Sultan Mfalme Hasan bin Sultan Ibrahimbin Sultan Yusuf the Shirazi Alawi (Sulte Mufalome Ase bone Sultane EbrahimoBone Sultame Uxufe Sirasice Lauy), send this letter to the Governor ofMorambique, wishing him good health, and also apprising him of my own. I alsosend my ambassador, Musa Muhammad, of the Malindane caste (MuxaAmamhamad Casta Melindane), whom I have ordered to seek news ofPortugaland to learn about former assistance, because all the people of my Kingdom-Mombasa, Mafia, Mongalo, and Zanzibar-are ready to receive Portugal's aidagainst the Arabs, so that theymight restore these lands to the Portuguese.The reply to this note states only that it is impossible to respond to the sultan'srequests until the annual fleet from Portugal arrives in August.On the other hand,it gives a different version of both the ruler's name and that of his ambassador.The former is addressed as the King of Kilwa, Sultan Hasan b. SultanIbrahim b.Sultan Muhammad b. Sultan YusufKilwi(:) Shirazi (Sultane AssaneRey deQuiloa, Ebuno Sultane Ebraimo, Buno Sultane Momade; Buno SultaneEusufo,Quiloae sirazi), while his envoy is referred to as Mwenyi Musa MwenyiMuhammad, "a Moor of the Malindane caste" (Muenhe Chaa! casta mouroMelindane Moenhe Mumade).7Comparing the two different versions of the sultan's name with thetext of the1776 treaty concluded by the same ruler with Morice, it seems likely that thediscrepancies can be resolved. The clerk who translated the sultan's letterapparently dropped Muhammad, the penultimate name in the pedigree, while thescribe who drafted the reply seems to have deleted the last name inthe line,Alawi, and reversed those of the sultan's grandfather and great-grandfather, Yusufand Muhammad. If we accept this reconstruction, we have fair evidence that theSultan of Kilwa in 1771 was the same Sultan Hasan b. Ibrahim b. Yusufb.Muhammad b. Alawi who was reigning at the time of Morice's first visit, somefive years later.6. A.N.F., Colonies s6rie C4, vol. 29, "Extrait d'une lettre ecrittede mozambiqueau J. Brayer du Barre en forme d'instruction et yjoint une lettre ecritte en arabe parle Roy de quilloa," n.d., but between extracts dated 15 March and 8 October,1770. Although a letter in Arabic script follows these extracts, it is neither inArabic nor from the Sultan of Kilwa, but rather "in ancient Lamu Swahili inArabic script from Alawi b. Sayyid Ahmad b. Abdallah to his brother MwinyiWaziri Abdallah bin Seyyidina Ahmad b. Abdallah, asking him to lookafter hiswife and to bring her to Pate." The letter is undated. (Freeman-Grenville, 1965b,p. 11.).7. A.H.U., Mogambique, caixa 14, translation of letter from King ofKilwa, n.d.,and reply (in a different hand), Mogambique, 12 April, 1771.

Edward Alpers

This fact throws some interesting light on what we know about the history ofKilwa at this time. According to Morice (Freeman-Grenville, 1965a, pp. 151-152), during the first three months of each year, the Arabs withdrew to Zanzibarin order to purchase their stock in trade from the Omani merchant vessels. Eventhe Arab governor of Kilwa joined this annual exodus, thereby temporarilyleaving the island free from his control. On the occasion of this phenomenon in1771, with the Arabs safely away in Zanzibar, the people of Kilwa simplydeclared their independence and notified the Arabs that they would infuture betolerated only as traders. Morice goes on to say that the Arabs believed the peopleof Kilwa to be backed "by a fairly strong party because they sent noarmed men.Note that the people of Kilwa were supported by the people of Mafia and theadjacent islands and by the whole Kingdom of Kilwa." This description accordswell with Sultan Hasan b. Ibrahim's contemporary picture of generaldissatisfaction with Omani overrule along the entire coast, even if the Portuguesetranslation of his letter gives a grossly inflated representation of the extent of hisdomain. Furthermore, although there is in Morice's sketch no indication of thesultan's role in this quiet revolution, his several attempts to enlist theaid of theFrench and the Portuguese (although their help proved unnecessary) stronglysuggests that he was at the very least the rallying point for the Kilwapatriots whoengineered the rebellion. It also seems more probable that an internally dynamicmovement of this nature would have taken place in these circumstances, ratherthan during the reign of either a dullard or a drunkard (cf. ibid., p. 41).A final point of interest which emerges from Sultan Hasan's last endeavour toensure Portuguese support for his cause (which may have been conveyed at thevery moment of Kilwa's ultimatum to the Arabs) is that his ambassador wascertainly the same individual who is named in the 1776 treaty with Morice asBwana Mwinyi Musa Malindane. As Secretary to the Sultan it was through hisoffices that Morice was bound to buy slaves, "in order to avoid all difficulty atKilwa" (ibid., p. 70). Hitherto, Morice's several references tothe office ofMalindane have been unique, and Freeman-Grenville presents an interestinganalysis of the problem in his introduction to the text (ibid., pp. 42-47). But inview of these new notices of the Malindane, the conclusions he draws about thetitle itself may possibly require re-examination.A brief recapitulation of the firmly established dates which we now possess seemsin order at this point. In 1723 Sultan Ibrahim b. Yusuf was reigningat Kilwa. By1759 he had been succeeded by his son, Hasan b. Ibrahim, who was still Sultan ofKilwa when Morice concluded his treaty in 1776. Have we any reasonable basesfor speculation regarding the regnal dates of these two sultans beyond these facts?Certainly a few exist, but they must be regarded with care.In January 1719 the Portuguese Viceroy of the State of India addressed a letter ofamity to the Sultan of Kilwa, noting the oppressive behaviour of the Arabs thereand sending him a small gift in token of Portuguese friendship.8 The letter andgift were apparently delivered secretly, to avoid arousing the Arabs' suspicion. Aswas to happen half a century later, however, Kilwa did not need external aid inorder to expel their Omani overlords, for after two decades of foreign rule, the

people of Kilwa reasserted their ancient independence early in the 1720s(Strandes, pp. 275-278). The next bit of8. British Museum, Additional Ms. 20, 906, f. 211, D. Luis de Meneses, Conde deEriceira, to King of Kilwa, Goa, 27 January 1719.

156 Sultans of Kilwainformation we have about Kilwa concerns the death of Fr. Francisco dePrezentago, also known as Francisco da Costa, who is reported to havebeenkilled "by the Mouros Monjungulos in Kilwa on the occasion of the conquest ofMombasa in 1729." (Silva Rego, p. 675). This is a perplexing reference. On theone hand, the two standard authorities on the Portuguese re-occupation and finalloss of Mombasa in 1727-1729 nowhere mention the presence of Portuguese atKilwa (Boxer, pp. 75-81; Strandes, pp. 278-294). This does not, of course, ruleout the possibility that a single Portuguese friar, perhaps encouraged by theeviction of the Arabs earlier in the decade, was not ministering his faith at Kilwathereafter. On the other hand, it is difficult to make sense of the phrase MourosMonjungulos. Presuming Monjungulos to be a variant of Musungulos, the usualPortuguese name for the Nyika living on the mainland behind Mombasa (seeibid., p. 351), we might read this as "Swahili(zed) Nyika." If this reading iscorrect, we must explain their presence at Kilwa, although this is not out of thequestion, provided that they were visiting sailors, or merchants, from Mombasa.But it seems more likely that the person who recorded Fr. Francisco's death, ifindeed it took place at Kilwa, knew very little about Kilwa and the exactcircumstances of this event, except perhaps that the assassins were AfricanMuslims. He may therefore have assumed that they were Islamized Nyika, whoprobably would have been a type not unfamiliar to the Portuguese inthe East.The next of these speculatory references is less open to such interpretation, but isalso less likely to be rewarding. In the late 1730s, during the administration of thecorrupt Nicolau Tolentino de Almeida, a son of the Sultan of Kilwa engaged inabortive negotiations at Mogambique with the governor for the establishment of"a new Trade" between the two towns.9 A few years later, in 1742, there was anexchange of letters between the new governor and the same Sultan ofKilwa, whoagain sent his son to act as ambassador to the Portuguese. Unfortunately, althoughthe Portuguese correspondence refers to letters from the sultan and from otherSwahili from north of Cape Delgado, these appear now to be lost.10 It is temptingto identify this sultan as Ibrahim b. Yusuf, and his son as the future Sultan Hasanb. Ibrahim. This cannot be done positively, however, although it does seem morelikely that the Sultan of Kilwa was at this date the father, rather than the son.The last of these sources, though much later in date, may help us to understandthe intriguing notice of the death of Fr. Francisco da Prezentarao. Late in 1784, acertain French slaver named Joseph Crassons de Medeuil rather hastily attemptedto negotiate a treaty of protection with the Sultan of Kilwa. Crassonsclaims that aletter to this effect was drafted to the King of France, stating among its provisionsthat in return for his protection "they offer to hand over a part ofthe island on itsNorth Coast in which is situated the fort formerly belonging to the Portuguese[the Gereza] and from which they were driven out by the present king's father.. ."

(Freeman-Grenville, 1962b, p. 194). Whatever the provenance of this last piece ofinformation concerning the father of the reigning sultan, there seems to be a realpossibility that Crassons' evidence makes sense.As I see it, there are four possible interpretations of Crassons' statement. One isthat Sultan Hasan b. Ibrahim was still reigning in 1784 and that his father wasalready sultan9. A.H.U., Movambique, caixa 3, Almeida to Crown, n.d., but probably c. 1741-1742. 10. Centro de Estudos Hist6ricos Ultramarinos, Lisboa, Filmoteca, Livrodas Monq6es 115, fl. 183 and 170171, jos6 Gongalves Valverde to D.Lourengode Noronha, Cape Delgado Islands, 20 June 1742, and Noronha to Viceroy,Mogambique, 14 August, 1742. The latter is published in Braganga Pereira, p.256.

Edward Alperswhen the Portuguese were driven from the Kenyan and Tanzanian coasts,including Kilwa, in 1698. A second is that either there were other Portuguese inaddition to Fr. Francisco at Kilwa in 1729, or that in view of their finalwithdrawal from Mombasa in that year, the concurrent murder of asinglePortuguese priest at Kilwa constituted a second expulsion of the Portuguese in thefolklore of the town some six decades later. In either case, it is virtually certainthat Sultan Ibrahim b. Yusuf was reigning in 1729. A third interpretation is thatthe driving out of the Arabs in 1721 and the murder of Fr. Francisco in 1729 bothoccurred in the reign of Ibrahim b. Yusuf, and that the two events hadbeensynthesised in local memory by 1784. In both these interpretations, Hasan b.Ibrahim remains as sultan at the time of Crassons' visit. Finally, one could arguethat the expulsion of the Arabs in 1771 from their headquarters in the Portuguese-built Gereza (the Omani reconstruction dates from after the reoccupation of 1785:see Gray, 1964, p. 24; Freeman-Grenville, 1965a, pp. 53-54, 56; Strandes, p. 338)was somehow falsely conveyed to, or misunderstood by Crassons as meaning thatthe Portuguese were thrown out during the reign of Sultan Hasan b. Ibrahim. Inthis case, it would have been his son, presumably Abu Bakr, with whom Crassonsmade his bargain.Of these four readings, I think that the last can be rejected straightaway, as itcarries far less authority than the others. My own preference has hitherto been forthe conclusion indicated by both the second and third interpretations of thematerial, although I would discard the totally speculative idea of a greaterPortuguese presence at Kilwa in 1729. Nevertheless, the first possibility isrecommended by our knowledge that the Portuguese certainly were present insome strength at Kilwa until the end of the seventeenth century (seeAlpers, 1966,pp. 56-59), despite the fact that it implies an unusually long regnalperiod (at leasteighty-six years) for only two sultans, father and son. Furthermore, although thegenealogical details are rather confused in the Kisiwani tradition (see above, p.146), there is good reason to believe that Sultan Ibrahim b. Yusuf succeeded as aminor, and that his great-aunt (not his aunt, as the chronicle states), Fatima bintiAli, acted as regent until he came of age. (See Freeman-Grenville,1962b, pp.222-223, and cf. his analysis of this passage in 1965a, p. 35). Even if he were in

his early twenties in 1698, he would only have been in his sixties in 1742, when itvery likely seems he was ruling. That the Habari za Zamani za Kilwa Kisiwaninames Hasan as the elder of Ibrahim b. Yusuf's two children doesnot seem to be aserious problem, either; this may merely be a way of indicating that hewas theelder of the two sons who succeeded to the title. In any case, Morice believed thatHasan was the youngest of three brothers. Considering the available sources, then,I think that this first interpretation is tentatively to be recommended.But withoutfurther documentation of a more positive nature, we can do no more thanspeculate on this point. What seems very probable, however, is the conclusionthat Sultan Hasan b. Ibrahim was reigning in 1784, and that it was hewhonegotiated with Crassons, as he had with Morice less than a decade before.Of the remaining new documentation there is nothing which adds to ourknowledge of the enigmatic Sultan Abu Bakr b. Hasan b. Ibrahim. All Ihavefound is a rather typically frustrating Portuguese reference in 1794 to the King ofKilwa, "whose letter I am sending enclosed to Your Excellency." Butlike somany other enclosures, this one is not to be found in any of the correspondencefrom Mogambique.11 There is, however, a good bit11. A.H.U., Morambique, caixa 28, Manuel Ant6nio Correia to D. Diogo deSousa, Ibo, 22 February 1794.

1h8 Sultans of Kilwaof valuable information on Sultan Yusuf b. Hasan.We have already noted that Prior did not record this sultan's patronymic in 1812(see above, p. 148). At the beginning of the following year, just four months afterthe Nisus left Kilwa, however, a brief letter of introduction was received atMorambique by the governor from Sultan Yusuf b. Sultan Hasan b. SultanIbrahim (Sultane Osufo Buno Sultane Assane Buno sultane Ibraimo). Although itscontents are inconsequential, noting only that an unnamed brother, coming fromZanzibar, wishes to trade at Mogambique, this letter clearly establishes the placeof Prior's Sultan Yousou Fou in the genealogy of his forebears.12On 29 March,1817 a letter of rather more importance was addressed from Mozambique by thenew governor to the same Sultan Yusuf b. Sultan Hasan (Sultao Sufu SultioAssane). Jos6 Francisco de Paula Cavalcanti de Albuquerque, the governor, waswriting in reply to an otherwise unknown letter from Sultan Yusuf to hispredecessor, whom he had replaced less than two months before (Bordalo, p.126). Recapitulating the contents of the sultan's letter, he notes that"you informedhim (the previous governor) of having thrown the Sakalava who had been thereout of Kilwa in less than three days."1 Here we have documentary substantiationfor an important event which is clearly recalled in the Habari za Zamani za KilwaKisiwani as being towards the end of Yusuf b. Hasan's reign (seeabove, p. 149).It should be noted, too, that this raid on Kilwa was previous to the more famousSakalava expedition against the Mafia Islands, which took place a year later inlate 1817 and early 1818. Indeed, the Kilwa Kisiwani tradition makes acleardistinction between these two raids. Given the date of Cavalcanti deAlbuquerque's letter to Yusuf b. Hasan, and in view of our knowledge of theorganization of the Sakalava raids (see Alpers, 1966, pp. 248-251), it seems most

likely that the raid against Kilwa took place towards the end of 1816, orat thevery beginning of 1817. This accords well with the invaluable observations of theFrench traveller Frangois-Joachin-Fortun6 Albrand, who noted in 1819 that"Kilwa was attacked not long ago by the Sakalava... [who] last yearformed anexpedition against Mafia..." (Albrand, p. 82).Fortun6 Albrand was a gifted young Oriental linguist who also had compiled awellreceived Arabic dictionary of 25,000 words before he left France in 1817 forR6union (which was still at that time called Bourbon), where he combined ateaching post with the duties of secretary to the administrateur g~ngral of thecolony. Two years later the Governor of Bourbon sent him as ambassador to theSultan of Kilwa, who is said to have proposed to cede Zanzibar and Mafia toFrance. As we have seen, there was a well established tradition for this sort ofploy by the Sultans of Kilwa. Accordingly, Albrand set sail aboard the corvetteAmaranthe on 16 January, 1819. (cf. Gray, 1964, p. 22, n. 2.) Unfortunately, wedo not know the duration of Albrand's sojourn in East Africa, although he spentenough time at Kilwa and Zanzibar to learn some Swahili, an achievement whichsets him apart from both Morice and Crassons. It appears, however, that Albrandreturned to Bourbon in the same year (Roman d'Amat, p. 1288).Albrand's remarks on the Sakalava raids constitute only one half of his validationof the testimony of the Kisiwani chronicle. Those on the sultanate itself (pp. 82-83) not only furnish us with useful supplementary information on the mode ofsuccession to the 12. A.H.U., Mogambique, caixa 57, Sultan Yusuf b. SultanHasan b. Sultan lbraim to Governor of Mogambique, Kilwa, 17 February, 1813.13. A.H.U., C6dice 1377, fl. 210.

Edward Alpersthrone (cf. Freeman-Grenville, 1962a, p. 151; 1965a, pp. 45, 81), but also confirmand add to the account of Sultan Yusuf b. Hasan's death, and the struggle whichfollowed it, in the Swahili history:The kingdom of Kilwa is at the same time hereditary andelective. The crown cannot leave the reigning family, but all the relatives of thesame degree of the deceased (ruler) have equal right to it, and thechoice must bemade among them by thedeputies of the diverse tribes of the coast.At the time of our arrival at Kilwa, the last king had just died,and his successor was not yet elected. Two brothers of the deceased aspired toreplace him, and this rivalry seemed to excite no misunderstanding among them.The elder of these princes, named Sulaiman (Soleiman), has a lively andenterprising character; his animosity against the Arabs is extreme, but he knowsto concealit.Thus we have positive evidence that Sultan Yusufb. Hasan died sometime beforeAlbrand arrived at Kilwa, probably either at the very end of 1818 orin the earlymonths of 1819. This dating also substantiates the implication in the Habari zaZamani za Kilwa Kisiwani that he died shortly after the last Sakalava raid on theTanzanian coast. Thanks to Albrand we now also know that Sultan Sulaiman b.

Yusuf was older than his brother Muhammad. Nor, being an outsider, is itsurprising that Albrand should not have been privy to the apparently bitter rivalrybetween the two brothers, which is so vividly related in the Swahili tradition.Finally, Albrand also mentions that a nephew of the Sultan reigningat the time ofthe final Omani occupation of Kilwa, early in 1785, distinguished himself in thefighting against the Arabs. This man, Mfalme Hasan, cannot be identified withany other known member of the royal family.Two further casual references at this time to undiscovered correspondence fromthe Sultan are no help at all, although the earlier must relate to the reign of Yusufb. Hasan, while the latter might conceivably indicate that his son, Muhammad,had actually succeeded by October 1819.14 On the other hand, a brief note ofamity from the Governor of Mogambique in 1830 establishes that the reigningmonarch was a Sultan Hasan (Assane), who was continuing the old tradition ofmaintaining official contact with the Portuguese through the officesof acommercial ambassador. As often before, this man was a relative ofthe sultan, inthis case a nephew named Husain b. Umar (Usene Bon Omar).15 Neither he norhis father are mentioned in the traditions. Despite the absence of this sultan'spatronymic, I think that we can safely identify him with Sultan Hasan b.Sulaiman, who is identified in the Habari za Zamani za Kilwa Kisiwani as the lastSultan of Kilwa. It is most likely, as the chronicle tells us, that he wassultan inname only; indeed, his lack of influence may well account for what appears to behis reasonably lengthy tenure of office. (cf. the Kivinje chronicle above, p. 152).Only the renewed threat in 1842 of 14. A.H.U., C6dice 1380, fl. 203-204,Cavalcanti de Albuquerque to Tomis Ant6nio de Vilanova Portugal, Morambique,9 March, 1818; A.H.U., C6dice 1394, fl. 14, Jolo da Costa Brito Sanches toConde dos Arcos, Morambique 10 October, 1819.15. A.H.U. C6dice 1425, fl. 4, Paulo Jos6 Miguel de Brito to Sultan Hasan ofKilwa, Mogambique, 13 March 1830.

Sultans of Kilwaresence at Kilwa moved Sayyid Sa'id to take the final step and deport Sultanuscat.written at some length on the present subject not so much because of itsimportance, but because of the methodological problems which are raised by cuesources involved. As I hope I have demonstrated, these include not only thetraditional histories, but also the documentary evidence which wehave examined.It is only by the careful and attentive use of both these sorts of materials, wherethe latter is available, that we can hope to achieve any accuracy in reconstructingthe chronologies of African polities, both coastal and interior.After this article was submitted for publication, I was enabled to carry out aprogramme of research in the Historical Archives of Goa, India, through thegenerosity of a grant from the Committee on Research of the University ofCalifornia, Los Angeles. Foremost among the documents which I collected werefourteen letters which have survived in the original Swahili texts, written inArabic script, as well as in the contemporary Portuguese translations. Theseletters span the second and third decades of the eighteenth century and are

concerned with the attitude of the Swahili rulers to their new Omani masters.Important as this new material is, for the present I wish only to note briefly thosedetails which bear upon the present article.In the first place, my previous suggestion that Sultan Ibrahim b.Yusuf was rulingin 1698 is now proved to be too generous, while the discovery of the relevantdocumentation is a clear reminder of the limitations of the sort of reconstructionwhich led me to that conclusion. In 1711 the Queen of Kilwa wrote two letters toone Mwinyi Juma b. Mwinyi Kwaja, a Mombasan resident of the Kerimba Islandswho had served as a Portuguese spy to the northern coast in the previous year (seeStrandes, pp. 275-276, where he goes unnamed). Although the Queen remainsanonymous in the Portuguese translations, the Swahili texts of the letters nameher as Mfalme Fatima.16 At the same time, Juma b. Kwaja received a joint letterfrom two sons "of the deceased King of Kilwa" whose names are given asMuhammad b. Yusuf (Mohamed buni euSuf) and Ibrahim b. Yusuf (Ibrahimobuni eusuf). A covering letter from the "Portuguese" spy refersto Muhammad as"Prince", while another clearly indicates that he was a more important figure atKilwa in 1710-1711 than his brother Ibrahim.17 So we now have positiveevidence that Ibrahim b. Yusuf was not sultan until after 1711, andthat anMfalme Fatima, who must certainly be identified as the Fatima binti 'Ali oftradition, was ruling at that date, apparently as regent for Muhammad b. Yusuf.This evidence would also seem to validate my placing of this unique female rulerof Kilwa in the royal genealogy. Incidental to this more significant material is theappearance of a daughter of Mfalme Fatima whose name is given as MwanaNakisa (Mana naquissa).18 16. Historical Archives of Goa, Livro das Mon6es 77,fl. 98-99, 101r, 102r. 17. Ibid., fl. 95r, Portuguese translationdated Goa, 25September 1711; fl. 90r "Informagao g trouxe MuenheJuma.....", n.d.; fl. 87-88,Juma b. Kwaya to Viceroy, Moganbique, 15 August 1711.18. Ibid., fl. 104r, Mwana Nakisa to Juma b. Kwaya, translation dated Goa, 25September 1711.

Edward AlpersThe only other information which bears on this problem concerns thereign ofSultan Ibrahim b. Yusuf. There is, unfortunately, no indication of his apparentlysenior brother, Muhammad, but we now have enough evidence to conclude thatIbrahim had assumed the throne by 1720, at the very latest. In 1720 and 1721 acertain Mwinyi Wali(?) Muhammad (Volay Mamede) carried out two separate missions to theSultanof Kilwa. Like his earlier counterpart, Juma b. Kwaja, he was also an inhabitant ofthe Kerimba Islands. He seems to have been entrusted with the Viceroy's letter of1719 to the Sultan, which I mentioned above on page 155. Following thecompletion of his two missions Muhammad was given a letter of recommendationto the King of Portugal from the Sultan of Kilwa who received him on bothoccasions, Sultan Ibrahim b. Sultan Yusuf b. Sultan Muhammad b. Sultan Alawi(Sultani Ibraimo, Buno Sulty. OSufo, Buno Sultany Musmady Buno SultaniAnby). Although there is some confusion in the Portuguese rendering of thenames of his grandfather and great-grandfather (only the translation of this

particular letter survives), our previous knowledge of Sultan Ibrahim'spatronymics facilitates the reading of this garbled version.19 It also seems verylikely that this is the letter for which I searched in vain in Lisbon (see above, p.145). Finally, at the other end of his reign, we now have a letter written by thissame sultan in about 1729.0The diagram overleaf attempts to set out the chronology of the sultanate of Kilwain the period considered in this article.19. H. A. G., Livro das Mon46es 93-A, fl. 4, King of Kilwa to Crown, n.d., butante-1724. 20. H. A. G., Livro das Mon4 es 97-B, fl. 617r, MfalmeIbrahim b.Sultan Yusuf(Falame Ibrahimo buno Sultane Suffo), King of Kilwa, to Governorof Mogambique, Ant6nio Cardim Frois, n.d., but c. 1729-1732. Theimprecisionof the dating of the documents in this section derives from the fact thatI amnecessarily relying on the Portuguese translations, which do notgive the dates ofthe original letters. This problem should eventually be eliminated, however, whenmy colleagues at the Institute of Swahili Research complete their examination ofthe photocopies of these fourteen precious documents.REFERENCESAbdallah b. Hemedi 'lAjjemy 1963 The Kilindi, ed. J. W. T. Allen andWilliamKimweri, Nairobi: East African Literature Bureau, in associationwith the AfricanStudies Program, Boston University.Albrand, F. 1838 "Extrait d'un M~moire sur Zanzibar et sur Quiloa,"Bulletin de laSocidti de Giographie, Paris, 2e S~rie, No. 10, pp. 65-84. Alpers,E. A.1965 Review of G. S. P. Freeman-Grenville, The French at Kilwa IslandAnEpisode in Eighteenth-century East African History, in Journal of AfricanHistory, Vol. VI, No. 3, pp. 418-419.1966 "The role of the Yao in the development of trade in East-Central Africa,1698-c. 1850," Ph.D. Thesis, University of London. Bordalo, F.M. 1859Ensaios sobre a Estatistica das Possess7es Portuguezas no Ultramar, Serie II,Livro IV, "Provincia de Mocambique," Lisboa: Imprensa Nacional. Boxer, C. R.1960 "The Portuguese on the Swahili Coast, 1593-1729," in Boxer andC. de Azevedo, Fort Jesus and the Portuguese in Mombasa, 1593-1729, London:Hollis and Carter, pp. 11-86.Bragania Pereira, 1938 Arquivo Portugues Oriental, nova ediggo,Vol. iv,Tomo 2, Parte 2.A. B. de (ed.) Bastori-Goa.

Sultans of KilwaBurton, R. F. Chittick, H. N. Coupland, R. Freeman-Grenville, G. S.P.Gray, J. M.Hoppe, F. Prior, J. Roman D'Amat Silva Rego, A. da Strandes, J.TeixeiraBotelho, J. J.Velten, C.1872 Zanzibar: City, Island, and Coast, London: Tinsley Brothers, 2,vols.1965 "The 'Shirazi' Colonization of East Africa," Journal of African History,

Vol. VI, No. 3, pp. 275-294.1966 "Kilwa: A Preliminary Report," Azania, Vol. I, pp. 1-36. 1938East Africaand its Invaders from the earliest times to the death of SeyyidSaid in 1856, Oxford: Clarendon Press.1962a The Medieval History of the Coast of Tanganyika, Oxford: ClarendonPress.1962b The East African Coast-Select Documents from the first to the earliernineteenth century, Oxford: Clarendon Press.1965a The French at Kilwa Island-An Episode in Eighteenth-century EastAfrican History, Oxford: Clarendon Press.1965b "Some Eighteenth-century Documents concerning Eastern Africain the Archives de France", Draft paper.1966 "Chronology from Genealogical Evidence: The East African Coast,"paper presented to the School of Oriental and African Studies,University of London, Conference on African Chronology.1962a "The French at Kilwa in 1797," Tanganyika Notes and Records,Nos. 58/59, pp. 172-173.1962b History of Zanzibar from the Middle Ages to 1856, London: OxfordUniversity Press.1963 "Zanzibar and the Coastal Belt, 1840-1884," in R. Oliver andG. Mathew (eds.), History of East Africa, Vol. I, Oxford: ClarendonPress, pp. 212-251.1964 "The Recovery of Kilwa by the Arabs in 1785," Tanganyika Notesand Records, No. 62, pp. 20-26.1965 Portugiesisch-Ostafrika in der Zeit des Marquis de Pombal(1750-1777),Berlin: Colloquium Verlag, Bibliotheca Ibero-Americana, Band 7. 1819 Voyagealong the Eastern Coast of Africa, to Mosambique, Johanna,and Quiloa: ... in the Nisus Frigate, London: for Sir Richard Phillips&Co.1933 Article on Fortun6 Albrand in Dictionnaire de Biographie Franvaise,Vol. I, Paris, pp. 1288-1289.1955 Documentalro para a Histria das Miss6es do Padroado PortuguesdoOriente, Vol. XI, Lisboa: Agencia Geral do Ulramar.1961 The Portuguese Period in East Africa, trs. J. F. Wallwork, ed. J. S.Kirkman, Kenya Historical Society, Transactions, Vol. II, Nairobi:East African Literature Bureau.1934 Histdria Militar e Politica dos Portugueses em Movambique da descobertaa 1833, Lisboa: Centro Tipogrdfico Colonial.1907 Prosa und Poesie der Suaheli, Berlin: Im selbstverlag des Verfassers.

THE SULTANATE OF KILWA IN THE EIGHTEENTH AND NINETEENTHCENTURIESSulaimanSULTAN 'ALI/ALAWIII

FATIMA binti 'ALIClearly documented: ft. 1710-1711, as regent during minority ofMuhammad andIbrahim b. YusufIMwana Nakisa b. ?Clearly documented: ft. 1710-1711iSULTAN MUHAMMAD b. 'ALI/ALAWISULTAN YUSUF b. MUHAMMADISULTAN IBRAHIM b. YUSUFMuhammad b.YusufClearly documented: ft. 1710-1711, apparently as heir designateClearly documented:ft. 1710-1711 (in minority), reigning 1720-c. 1729;suggested,fl. post-1737 to 1742I I? b. Ibrahim Sultan Yusuf b. IbrahimClearly documented: Strongly inferred:ft.fl 1776 1776 (one of two deposedSultans met by Morice ?)II IMfalme Hasan b. ?Strongly inferred: foughtagainst Arabs at Kilwaearly in 1785ISULTANYUSUF b. HASANClearly documented: ft. 1813-1817; strongly inferred:ft. 1812; d. late 1818--early1819. Almost certainly named by Prior in 18121Mfumo Hasan b. YusufClearly documented:ft. 1812ISULTANABU BAKR b. HASANISULTANHASAN b. IBRAHIM *Clearly documented: ft. 1759-1776; strongly inferred: ft. 1777-1784iSULTANSULAIMAN b. HASANMwinyiJumaMwinyiJamoto (?) b. ?

Clearly documented: ambassador to Mozambique, 1759SultanMuhammad b. HasanSULTAN HASAN b. SULAIMAN ........... 'Umar b. ?Strongly inferred:ft. 1830; deported by Sayyid Sa'id c. 1842/1843Husain b. 'UmarClearly documented: ambassador to Mocambique, 1830* The Habari za Zamani za Kilwa Kisiwani says he is older than SultanYusuf b.Ibrahim. KEY: NAMES IN ALL CAPITAL LETTERS: documented SultansNames in Upper and Lower case Letters: documented members of Royal FamilyItalics: named only by tradition.

A z a n i aThe Journal of the British Instituteof History and Archaeology in East AfricaEdited byNEVILLE CHITTICKVOLUME II - 1967Published on behalf of theInstitute byOxford University PressNairobi AddisAbaba Lusaka1967

Oxford University Press, Ely House London W. IGLASGOW NBWYORK TORONTO MELBOURNE WELLINGTON CAPETOWN SALISBURY IBADAN NAIROBI LUSAKA ADDIS ABABABOMBAY CALCUTTA MADRAS KARACHI LAHORE DACCAKUALA LUMPUR HONG KONG TOKYO© British Institute of History andArchaeology in East Africa, 1967Made and printed in East Africa