Mafia Island.
----------------
The first mention of Mafia is in the Kitab Ghara'ib al-funun wa-mulah al-'uyun (1050) a place is mentioned: Manfiya, an island (Mafia) this according to Horton, M. (2018) in ‘The Swahili Corridor Revisited’ is the island of Mafia.
The second also as Manfia is Idrisi (1150). But unsure as it might be Mombasa.
Yakut (1220): Mnfyh is a town famous on the Zinj Sea coast.
Al Firuzabadi (1414) has Khna; maybe Kua on Juani Island (but unsure).
Kisimani
Taken from: The Archaeology of Islam in Subsaharan Africa by Timothy Insoll (2003)
On Mafia Island Kilwa coins, copper issued of Ali ibn al-Hassan, have been recovered at the site of Kisimani Mafia (Horton et al. 1986:121]. These coins, numbering over 600, were discovered in an imported sgraffiato jar of Persian Gulf origin. The Kilwa Chronicle records that the sons of Ali ibn al-Hasan settled at Mafia, probably at the site of Kisimani Mafia, according to Chittick (1961:1). This is a large site, a stone town, with accompanying wattle-and-daub areas as well. Chittick (1961, 1964) focused his attention initially on a stone mosque, in which five building phases dating from between 1240 and 1520 were discerned. The total site covered about 1ha. The remains of other stone structures have also been recorded at Kua on Juani, and a further mosque on Jibondo (Baumann 1957:16). More recently, Felix Chami has been working on Mafia, and his work will undoubtedly push back the occupation sequence, and place Mafia within the earlier tradition of Swahili centers as well.
Note about the Kilwa Chronicle: The founder of Kilwa, Ali ibn al-Husayn Ibn Ali, retired to Mafia in his old age, and set up his son Muhammad as ruler in ca. 996AD. He was succeeded by his brother, Basbat, who ruled for four-and-a-hall years, The third ruler of Kilwa, Dawud ibn Ali, also made Mafia his residence, after abdicating at Kilwa in favor of his son. In the reign of Dawid ibn Sulayman of Kilwa (ca. 1142-70AD), Mafia is mentioned as a mere appanage of Kilwa. It is not mentioned again until the reign of Sulayman ibn al Hasan II (ca. 1294-1308), when it revolted. It was recovered past 1310 by his son, al-Hasan ibn Sulayman II (ca. 1310-33).
Note2: Much of the headland, together with the town of Kisimani , was washed away by a cyclone in the eighteen hundreds. By now nearly all building remains are under water.
Taken from: Mafia Island Maritime and Underwater Cultural Heritage Survey 2012 (TMUCH)
In recording high and low water mark revealed that the building remains (of Kisimani Mafia) stretches for about 500 meters from west to east. Many buildings remains were located in the area, but also three wells.
Of the 31 copper coins recovered from the beach at Kisimani Mafia four were ‘foreign’ and the other 27 were smaller. The smaller coins, all similar in shape and size (16 mm diameter) were presumably minted at Kilwa Kisiwani from 11th to 14th/15th century as some show the names of the different Sultans of Kilwa Kisiwani which helped Chittick to date the Kisimani Mafia site.
Two of the foreign coins were Chinese, one identified as a kai yuan tong bao coin from the Tang Dynasty (618 to 907). The kai yuan tong bao coin is fully intact (24 mm in diameter) but very corroded, the other was only a fragment and it has not been identified.
Varieties of ceramic artifacts were recovered during the survey and a few were analyzed. Among these ten potsherds, 6 were local dating from Early Iron Age (EIA: 1st to 6th centuries) and four were imported dating to between 10th and 14th centuries.
Taken from: The Cultural History and Chronology of Kisimani Mafia Stone Town Site, Mafia Archipelago, Tanzania: Findings from Re-excavation of the Site. By F. Chami (2020)
This study involved an archaeological survey and excavations carried out in 2018 at Kisimani Mafia. An array of cultural materials was recovered, including pottery, beads, glass, coins, slag and metal objects, as well as bone and shell remains. This paper, however, is based on the results of pottery and chronometric analyses, the results of which firmly established that the stone town site was occupied from a much earlier period than was thought. The cultural history spans the Triangular Incised Ware/Tana Tradition period through the Plain Ware period, to the end of the Swahili Ware period. The first two periods, which evolved from the Early Iron Working culture of the archipelago, were when the Swahili stone town settlement was founded. (Starting third – fourth century).
Taken from: Kisimani Mafia; Excavations at an Islamic Settlement on the East African Coast; by H. N. Chittick 1961.
Kua on Juani island
There is an unsure mentioning as Khna by Al Firuzabadi (1414)
Taken from: Refining the chronology of Kua Ruins: radiocarbon dates from the Mafia Archipelago, Tanzania by Annalisa C. Christie
Situated on the southwest coast of Juani Island in the Mafia Archipelago, Tanzania, the settlement site of Kua is an extensive complex of stone structures covering an area of over 30 acres. The site comprises seven mosques, four cemetery areas, at least ten ‘complex structures’, approximately 30 “walled courts” (Garlake 1966: 109), which may have been associated with structures constructed from non-stone materials (Freeman-Grenville 1962: 193), and numerous internal and external latrines. Although the precise date of Kua’s foundation is unknown, it features on the maps of the Arab geographer al-Idrisi, which are thought to date to 1154 (Chittick and Rotberg 1975: 141).
Note: this identification with Kua on Mafia is uncertain. More recent authors state Kahua = Mayotte.
The original settlement is said to have been concentrated in an area to the south of the main ruins complex known
as Mkokotoni (literally ‘At the mangroves’). The popular narrative indicates that Kua was raided in the late eighteenth century by the Sakalava from Madagascar and subsequently abandoned.
The design of its mihrab suggests that the northernmost mosque dates to the fourteenth/fifteenth centuries AD (Freeman-Grenville 1962: 173; Garlake 1966: 69). The largest of the mosques recorded at Kua uses a similar decorative style, they were constructed at a similar time, with both Chittick (1957: 1) and Freeman-Grenville (1962: 173) suggesting a date of the fourteenth or fifteenth centuries AD. The other mosques at Kua, along with the majority of the other structures within the main complex, are thought to date to the seventeenth and eighteenth centuries (Garlake 1966: 108).
This would make them broadly contemporary with the only dated burial at the site, the inscription on which reads 1175 (AH), equivalent to AD 1761–1762.
Two types of Chinese pottery were recorded at Kua. The first was a single sherd of Chinese celadon ware, of the fourteenth/fifteenth centuries AD (Chittick 1957: 3). The second, more common type of Chinese pottery was blue-on-white porcelain, which Horton (1996: 16) indicates was ‘introduced to East Africa in the later 14th century’.
This plan of the ruins of Kua shows the ruins of the seventeenth and eighteenth centuries plus the two medieval mosques. Mkokotoni area (the oldest settlement area) is just south from this map.
Hinterland:
Taken from: The Early Iron Age on Mafia Island and its relationship with the mainland (1999) F. A. Chami.
One striking archaeological contrast between Mafia and both Zanzibar and Pemba islands, as well as the mainland opposite the Zanzibar channel and south as far as the Rufiji, is the virtual absence on Mafia of sites with pottery of the TIW (Triangular incised Ware) tradition.
On present evidence, it is possible that, after several centuries of Early Iron Age settlement, Mafia was deserted around the sixth century and not reinhabited until the eleventh, the latter being the time of intensification of maritime trade on the southern Swahili coast and of the rise of the Kilwa sultanate with which Mafia was closely associated.
Taken from: Overview of Work Conducted in the Mafia Archipelago 2008-2010 by Annalisa C. Christie
She concluded: Similar to other surveys conducted (e.g., Chami 1994, 2000, 2004; Wynne-Jones 2006) most of the sites recorded during the surface surveys and excavations were suggestive of more transient occupation – with shallow and dispersed artefact deposition. This suggests that Kisimani Mafia and Kua may have been the only stone town settlements on the islands. This raises the question of where (if at all) the others were located and, if these were the only two large settlements, why were others not constructed?
Note: In her excavations in six sites along the west coast no imports datable to the Middle Ages were found but on Chole island and close to Kua on Juani were Chinese sherds of 14th -15th century discovered.
Note: Conclusions; The remains from the Middle Ages so far have only been found in the areas close to Kilwa. Except of Kisimani and Kua there were also two other small settlements but in the same area. Northeast on Juani island. Walls and mounds of houses were seen covered by thick bush. Pottery from the nearby ruined mosque suggests that the site should be assigned to the Kua/Kilwa time period. (Felix A. Chami 2000) And also on Jibondo island at the coast are the remains of a 15th century mosque.
The only reason so far to explain this is that not enough settlements were existing on Mafia Island to make a solid trading hinterland.