Contemporary sources furnish fairly ample information on the structure of the Mamluk army and the units from which it was composed; but, though some of their definitions and descriptions come near to the truth, the present writer has found none of them to be completely accurate. He is of the opinion that the treatment of this question by modern Orientalists is also in need of some emendation, and he here submits the composition of the Mamluk army which, in his opinion, is the correct one, and the grounds for his suggestion will be presented in the course of the discussion.

The Türk ve Islam Müzesi in Istanbul possesses a hitherto unpublished ewer (Inv. No. 217) signed by a Mawṣilī artist (Pl. I). This is an unpretentious vessel, made of beaten brass and sparingly decorated with inlaid designs. It stands 37 cm. high without the lid (which is a later addition) and measures 20 cm. in its widest diameter.

In the year 577/1181–2, in a letter to the Caliph in Baghdad explaining his activities in Syria, Saladin writes that he is engaged in a struggle for Islam against a three-fold enemy—the infidel Frankish invader, the heretical and murderous Assassins, and the treacherous Zangid rulers of Mosul, whom he accuses of intelligence and even alliance with both Franks and Assassins.

The significance of Political Philosophy for the philosophy of Islam and A Judaism in the Middle Ages has been recognized in recent years, and the importance of Plato's Republic and Laws together with Aristotle's Nicomachean Ethics for the falāsifa is becoming increasingly evident. Attention has been drawn to Alfārabī's political treatises, to Ibn Bājja, to Averroes' Commentaries on Plato's Republic and on Aristotle's Nicomachean Ethics, to Ibn al-Ṭiqṭaqa and to Ibn Khaldun. The reception of Platonic and Aristotelian political ideas conditions—in varying degrees—the attitude of the falāsifa to the State as such, to the Muslim State in particular, and therefore to the Sharī‘a of Islam and to Greek philosophy, both practical and theoretical. This is already clear from the writings of Alfārābī, but in his attempt at blending and harmonizing Greek-Hellenistic and Islamic conceptions no Alfarabian philosophy proper with distinct features of a definitive, authoritatively formulated synthesis emerges. It was left to Averroes—to use the name by which Ibn Rushd was generally known to the West—to work out a peculiarly Islamic religious philosophy. He establishes in unequivocal terms the supreme authority of the Sharī‘a and its identity of purpose with falsafa, religious philosophy in the Faṣl al-maqāl, the Manāhij, the Tahāfut al-tahāfut and his Commentary on Plato's Republic. The last named forms the main subject of the present article. In an earlier study it was considered in connexion with his Commentary on Aristotle's Nicomachean Ethics already, but not with his other treatises just mentioned. This is, however, necessary in order to gain a correct, balanced view of his philosophy, and especially of his political thought. It is equally important to see his political Commentaries against the background of his active life as qāḍī and of contemporary Muslim history in the Maghreb.

The odes of Hafiz have been for some 600 years a source of delight to Persians, and an object of their serious study. To Europeans they have been known for less than half that time; yet here too, many scholars and poets have laboured to understand and interpret them. Much remains, no doubt, still to be done; but it is unlikely that anyone will now discover in the poems wholly new ranges of meaning, which have been hidden from generations of scholars among the poet's own countrymen. Any claim to such discovery must necessarily be subjected to the strictest scrutiny.

In this study of the population statistics of Ming China I have relied almost exclusively upon figures from contemporary sources: the Veritable Records (Shih-lu) of the dynasty, the Collected Statutes (Ta-Ming Hui-tien), the Imperial Topography (Ta-Ming Yi-t'ung-chih), the Huang-yü K'ao, and a number of provincial and local gazetteers.

The first question that arises in using such works is that of the reliability of their figures and indeed of Chinese statistics in general. A great deal of scepticism has been, and still is, expressed on this point. To take a single example only: everyone who has researched in this field has experienced over and over again that where a Chinese text gives both a global total and an itemized breakdown of that total, the sum of the separate items usually amounts to something more than (or less than) the global figure, and hardly ever agrees exactly with it. A natural reaction from this unsatisfactory state of affairs is to reject all the figures concerned as unreliable. To do this, however, is tantamount to denying the possibility of quantitative investigation in Chinese economic and institutional history. Before coming to such an extreme conclusion it is at least advisable to see whether the figures are capable of rational emendation, and whether figures so emended will show a higher degree of internal consistency.

‘Standard Chinese’ is here offered as a translation of Gwoyeu and is defined, for the purpose of this paper, as ‘polite Peking speech’, i.e. the language spoken by a youngish man, born and brought up in Peking, when he avoids any features which he himself would describe as tuuhuah or local—i.e. Peking—dialect. The grammatical features described are based exclusively on the speech system of Mr. James Liou, and may therefore include personal idiosyncrasies which other Gwoyeu speakers would not recognize as ‘standard’. Mr. Liou was born in Peking in 1927 and lived there until 1950. His family come from a village within 30 miles of Peking, and have lived in the capital for the last three generations. He has not therefore been subject to ‘non-Pekinese’ linguistic influences to any great extent.

Brian h. hodgson's vocabulary of the ‘Kirānti’ contains a jumble of dialects, those in adjoining columns seldom being closely related to each other. But his vocabulary has the advantage that all the dialects listed by him belong to one Tibeto-Burmic group, which we may call East Himalayish.

‘Particle’ and ‘Noun’ are not a priori categories of universal applicability. Ad hoc definition of their ‘meaning’ is given under A. Detailed description of structural relations observable between them follows at B.

It is recorded in a number of standard works on the western Sudan and northern Nigeria that about the year A.D. 1300 Sākūra, the usurping King of Mali, performed the pilgrimage to Mecca and was murdered by the Danakil on the return journey. Delafosse writes: ‘C'est en revenant des lieux-saints par le Yémen et l'Erythrée, vers l'an 1300, que Sakoura trouva la mort: il fut dévalisé et assassiné par des Danakil sur la côte de Tadjourah, comme il venait de débarquer sur la terre d'Afrique.’ Meek, de Pedrals, and Bovill give the same story, and the last-named comments: ‘He was not returning by the customary route through Cairo, but through Eritrea and the Eastern Sudan, following what has now become the great pilgrim road of West-Central Africa.’ Now if Sākūra was killed in what is to-day French Somaliland, or for that matter anywhere in Somali or Dankali country, he was far to the south of ‘the great pilgrim road’, which does not cross Eritrea at all. Nor is it conceivable that it should ever have done so. It goes through Bornu, Wadai, and Darfur to the Nile and now reaches the Red Sea at Port Sudan. There is no reason why pilgrims should have gone so far to the south of the latitude of Jidda and so have given themselves the task of climbing over the Abyssinian plateau which was, moreover, in Christian hands. A glance at the map will show what a very strange route Sākūra is alleged to have chosen. Having landed at Tajura he would have had three alternatives. The first was to follow approximately the line of the modern railway to Addis Ababa, climbing the steep escarpment in Shoa, and then making his way across Ethiopia and down the other side of the plateau to the Nile valley in Sennar, whence he could have reached Darfur. This would have been very difficult for topographical reasons and would have necessitated a long journey through Christian territory. The second possibility was to travel northwards along the coast until he reached a place such as Suakin where he could have turned westwards and headed for the Nile. This would have been even more difficult, given the character of the Danakil and their country. Last and most difficult of all, he could have attempted the little-used and circuitous route that led through the Sidama States, on the southern borders of Ethiopia, to the White Nile. It is almost inconceivable that he should have been trying to do any of these things.

Hardly had the article ‘A Farewell to the Khagan of the Aq-Aqatärān’ been printed, when my eye fell on a passage in Pliny which settles the problem posed by Mšyk. Shapur, it will be recalled, defeated the Roman army, under Gordianus (who fell in the battle), at Mšyk = MHCIXICH and MICIXH, and renamed the town (and district) Pērōz-Šabūr. The ancient name, in a form which precisely corresponds with the Parthian spelling Mšyk, is mentioned by Pliny, v, 21, 4 (§ 90): scinditur Euphrates a Zeugmate dlxxxxiv mp. circa vicum Massicen, et parte laeva in Mesopotamiam vadit per ipsam Seleuciam, circa eam praefluenti infusus Tigri; dexteriore autem alveo Babylonem … petit; a later passage, vi, 30, 3 (§ 120), indicates that the left branch referred to is the Narmalchas, the ‘Royal River’. The identity of Massice with Pēroz-Šabūr could have been recognized, even without the help of the inscription, by comparing Pliny with Ammianus Marcellinus, who made a statement in similar terms, xxiv, 2, 7, hinc pars fluminis scinditur … ducens ad tractus Babylonos interiores …, alia Naarmalcha nomine … Ctesiphonta praetermeat; the latter branch was crossed by the Roman army, which immediately came upon Pirisabora.

In S. H. Taqizadeh's article, BSOAS., xiv, p. 606, line 24, the first word should have been Ayāθrima in the place of Paitišhahya; and, pp. 610 sq., Nātanz is wrong for Natanz (Natanz). As I was responsible for seeing the article through the press, it is for me to apologize for these mistakes.