202 UMAIYAD PRACTICE AS THE STARTING-POINT

were his only property, but Abān drew lots and set free only the winning two.¹

This was projected back to the Prophet, first of all as a mursal, both in Iraq and in Hijaz, with the isnāds: Mālik—Yaḥyā b. Saʿīd and others—Ḥasan Baṣrī and Ibn Sīrīn—Prophet (Muw., ibid.), and Ibn Juraij—Qais b. Saʿd—Makḥūl—Ibn Musaiyib—Prophet (Ikh. 370). This tradition dates only from the second century, because Shāfiʿī states² that it is the only argument which can be adduced against the doctrine of Ṭāwūs on another problem of legacies; whether the alleged doctrine of Ṭāwūs is authentic or not, the tradition cannot have existed in the time of the historical Ṭāwūs who died in A.H. 101. The whole doctrine on legacies was still fluid at the beginning of the second century.

The restriction of legacies to one-third of the estate was the common ancient doctrine and was directly based on an Umaiyad administrative regulation. But as regards the manumission of slaves on the death-bed, the Iraqians, for systematic reasons, abandoned the drawing of lots and set one-third of each slave free (Ikh. 380 ff.).

For obvious fiscal reasons, the Umaiyad administration controlled the granting of ownerless and uncultivated land for purposes of cultivation.³ As far as the disposal of land already under cultivation, abandoned by its former owners at the time of the great conquests, is concerned, Muhammadan jurisprudence gives only an artificially systematized picture, which as a whole is considerably later than the facts it purports to represent. At the beginning of the second century A.H., when Muhammadan legal science began, there remained only the question whether a grant of the administration was necessary for a valid title to uncultivated land brought under cultivation for the first time (iḥyā' al-mawāt). Both the Iraqians, down to Abū Ḥanīfa, and the Medinese answered in the affirmative, upholding the Umaiyad administrative practice which in this case was maintained by the 'Abbāsids.⁴

In the generation preceding Mālik, however, traditions from—