(4) Ismaʿil, son of Jaʿfar,—Muhammad “the concealed,”—Ismaʿil,—Ahmad,—Ubayd Allah. This also occurs in Abu l-Feda, and in ʿUbayd Allah’s “Genealogy of the ʿAlids” (MS. Leiden, 686—cited by de Goeje, Qarmates, p. 9) Muhammad had three sons, Ismaʿil II, Jaʿfar, and Yahya; Ismaʿil had a son named Ahmad, who dwelt in the Maghrab.
(5) Ismaʿil—Muhammad “the concealed,”—Ismaʿil II,—Muhammad,—Ahmad,—ʿAbdullah,—Muhammad,—Husayn,—Ahmad or ʿAbdullah,—Ubayd Allah the Mahdi. This is the genealogy given in the sacred books of the Druses, and rests on the theory that there must have been seven “concealed Imams” intervening between Jaʿfar as-Sadiq and the Mahdi. It is merely an instance of the mystic value attached to the sacred numeral. Like (3) it gives Muhammad for Ahmad which is a permissible variant.
(6) The five preceding genealogies are distinctively Ismaʿilian in character, but there are others which show adaptations of the “Twelvers” accounts, and these cannot be much more than later attempts to connect the Fatimid line with that recognised by the other Shiʿites. First we have the idea that the descent from Jaʿfar as-Sadiq was through Musa, not Ismaʿil, then following the next three Imams ʿAli ar-Rida—Muhammad al-Jawad—ʿAli al-Hadi (see above)—al-Hasan al-Askari—Ubayd Allah the Mahdi. According to this the Fatimite Mahdi in Africa was the son of the eleventh Imam of the “Twelvers,” and thus replaced Muhammad al-Muntazar.
(7) The same line as the preceding, but admitting Muhammad al-Muntazar as twelfth Imam who “disappeared” in 260, and asserting that ʿUbayd Allah who appeared in North Africa was this same Muhammad emerging from concealment, after an interval of 29 years.
(8) The same line as far as ʿAli al-Hadi, then Husayn, presumably a brother of Hasan al-Askari, and ʿUbayd Allah as son of this Husayn. This is given by Ibn Khallikan on the authority of a reference in Ibn al-Athir. All these three last genealogies must be dismissed as later suggestions since it is clear that the Ismaʿilian sect rejected the Imams of the “Twelvers” after Jaʿfar as-Sadiq: but it may be that Ahmad’s first claim was simply to be an ʿAlid, and not necessarily the son of the house of Ismaʿil.
(9) Finally we have another theory, mentioned by Ibn Khallikan, that the Mahdi was descended from Hasan, a brother of Jaʿfar as-Sadiq, and so an ʿAlid but not an Imam, and from this Hasan came ʿAbdullah, Ahmad, Hasan, and then ʿAli or ʿUbayd Allah the Mahdi. Back to ʿAbdullah this was the generally asserted genealogy of the Mahdi’s family, but Hasan, the brother of Jaʿfar, replaces Maymun.
The chief point is that there were so many alternative forms of the genealogy, and close scrutiny shows very weak points in every one of them. To the fully initiated this was a very small matter, as no importance was attached to the claim to the Imamate or to the descent from ʿAli at all. No doubt all these pedigrees served their purpose in dealing with the different types of proselytes, and their very diversity tends to prove that they were actually accepted and circulated in a sect which adapted its teachings to suit the opinions of the different classes with which it came into contact. It was not until the Fatimids became a political power that any need was felt to bring these various genealogies into any kind of agreement, and then, no doubt, the variant forms circulated by the different missionaries were a source of embarrassment.