death.¹ One man would read a verse one way, and another man another way, each saying that his reading was better than his neighbour's, and contending for the same; and there was change and interpolation, some copies having more and some less. When this was represented to Othm�n, and the danger urged of division, strife, and apostacy, he thereupon caused to be collected together all the leaves and scraps that he was able, together with the copy that was written out at the first. But they did not interfere with that which was in the hands of Aly, or of those who followed his reading. Obey was dead by this time. As for Ibn Mas�d, they demanded his exemplar, but he refused to give it up, and so Abu M�sa was appointed governor of Kufa in his room.² Then they commanded Zeid ibn Th�bit, and with him Abdallah ibn Abb�s (others say Mohammed, son of Abu Bekr), to revise and correct the text, eliminating all that was corrupt. Now both were young;³ and they were instructed, when they differed on any reading, word, or name, to follow the dialect of the Coreish. On many points they did differ. For instance, Zeid wrote T�b�h, and Ibn Abb�s T�b�t. When the recension was completed, four exemplars were written

out in large text, and one sent to Mecca, and another to Medina. The third was despatched to Syria, and is to this day at Malatia (Melitene). The copy at Mecca remained there till the city was stormed by Abu Sar�ya (that is, the last time the Kaaba was sacked, A.H. 200); he did not carry it away ; but it is supposed to have been burned in the conflagration. The Medina exemplar was lost in the reign of terror, that is, in the days of Yez�d ibn Mu�via. The fourth exemplar was deposited in K�fa, then the centre of Islam and home of the Companions of the Prophet. People say that this copy is still extant there; but this is not the case, for it was lost in the insurrection of Mukht�r.¹

"After what we have related above, Othm�n called in all the former leaves and copies, and destroyed them, threatening those who held any portion back; and so only some scattered remains, concealed here and there, survived. Nothing remained to show the discrepancies which are known to have existed. It is said for example that Sura N�r (xxiv.) used to be longer than Sura Bacr (ii),² and that Sura Ahz�b (xxxiii.), is mutilated and incomplete; so also that there was originally no division between Sura Bar�t (ix.) and Sura Anf�l (viii.), and accordingly we see that the invariable heading. In the name of God most Merciful, is wanting in the former. Similar is the case of the two 'Incantation Suras,³ of which Ibn