A. J. Butler, The Arab Conquest of Egypt and the Last Thirty Years of the Roman Dominion, Oxford, 1902, Chapter 25, p.401 gives the following version of the story from Bar Hebraeus, or Abu’l Faraj, once again not quite in verbatim quotation:
The story as it stands in Abu’l Faraj  is well known, and runs as follows. There was at this time a man, who won high renown among the Muslims, named John the Grammarian. He was an Alexandrian, and apparendy had been a Coptic priest, but was deprived of his office owing to some heresy by a council of bishops held at Babylon. He lived to see the capture of Alexandria by the Arabs, and made the acquaintance of `Amr, whose clear and active mind was no less astonished than delighted with John’s intellectual acuteness and great learning. Emboldened by `Amr’s favour, John one day remarked, ‘You have examined the whole city, and have set your seal on every kind of valuable: I make no claim for aught that is useful to you, but things useless to you may be of service to us: ‘What are you thinking of?’ said `Amr. ‘The books of wisdom,’ said John, ‘which are in the imperial treasuries: ‘That,’ replied `Amr, ‘is a matter on which I can give no order without the authority of the Caliph: A letter accordingly was written, putting the question to Omar, who answered: ‘Touching the books you mention, if what is written in them agrees with the Book of God, they are not required: if it disagrees, they are not desired. Destroy them therefore: On receipt of this judgement, `Amr accordingly ordered the books to be distributed among the baths of Alexandria and used as fuel for heating: it took six months to consume them. ‘ Listen and wonder,’ adds the writer.
Such is the story as it makes its appearance in Arabic literature. Abu’l Faraj wrote in the latter half of the thirteenth century, and he says nothing about the source from which he derived the story : but he is followed by Abu’l Fida in the early fourteenth century, and later by Makrizi . I t is true that `Abd al Latif, who wrote about 1200, mentions incidentally the burning of the Library by Omar’s order, and, giving no details, seems to take the fact for granted. This allusion seems to show that in his day the tradition was current. …
Let us, however, examine the story as it stands. It is undeniably picturesque, and the reply of Omar has the true Oriental flavour. This really is the strongest point about it. But unfortunately precisely the same reply of Omar is recorded in connexion with the destruction of books in Persia …
 Ed. Pococke, p. 114 tr. and 180 text. Renaudot thinks the story has an element of untrustworthiness: Gibbon discusses it rather briefly and disbelieves it. Pococke translates only the Arabic abridgement of Abu’l Faraj. In the Nineteenth Century for October, 1894. there is an article on the question by Vasudeva Rau, who alleges (po 560) that the story is not in the original Syriac, and probably was a later interpolation.The abridgement, however, was written by Abu’l Faraj himself, and the suggestion of interpolation is a mere conjecture. Nor would the fact, if established, be material. The article generally is based rather on a priori argument than research, and consequently is not of much value.
 This author, like `Abd al Latif, reports the story by way of allusion, taking it for granted. Thus speaking of the Serapeum he says, ‘Some think that these columns upheld the Porch of Aristotle, who taught philosophy here: that it was a school of learning: and that it contained the library which was burnt by `Amr on the advice of the Caliph Omar’ (Khitat, vol. i. p. 159).
 See Prof. Bury’s ed. of Gibbon, vol. v. p. 454 n., where Ibn Khaldun, quoted byHaji Khalfah, is given as the authority. I may add that the feelings of the Muslims towards the books of the idolatrous Persians would be very different from their feelings towards the books of the Christians. In their early history at least the Muslims disliked the destruction of the written name of God.
This gives us the text of Makrizi, which is good, and yet another inaccurate translation of Pococke, which is not. I begin to think that what we need to do is collect the Arabic text of all the witnesses, and get them translated afresh and accurately.