¹ See my Marvazi, p. 147.

¹ Marvazi on China, the Turks, and India, Royal Asiatic Society, 1942, pp. 68 and 111.

² See Gibb, H. R. A., “ The Arab Invasion of Kashgharin 715 ”, BSOS., II/3, 1922,467–474; Gibb, “ Chinese Records cf the Arabs in Central Asia ”, II/4, 1922, 613–622; Gibb, The Arab Conquest of Central Asia, 1923, pp. 12–13Google Scholar

³ The ambassador of the Caliph Hisham (A.D. 724–743), on whom see Marquart in Festschrift fur F. Hirth, 1920, ed. Ostasiatische Zeitschrift, 289–293; the interpreter Sallam under Caliph Wathiq (A.D. 842–7), see Hudud al-'Alam, p. 225; Ibn Fadlan in 309/921, see the Russian translation by Kovalevsky, 1939, and the latest work by A. Z. Validi-Toghan, 1940 (a new edition with commentary is being prepared in the U.S.S.R.). On the spurious report of Abu- Dulaf's journey (supposed to have taken place circa A.D. 941) see Marquart, Streifziige, 74–95. Marquart finally came to the conclusion that the existing report was falsely attributed (untergeschieben) to Abu Pulaf, whose genuine work Marquart planned to trace separately, Festschrift f. E. Sachau, 1915, p. 292. The most recent work on Abu Dulaf is by A. von Rohr-Sauer: Des Abu Dulaf Bericht uber seine Reise nach Turkestan, China und Indien (Bonner Orientalische Studien), Stuttgart, 1939, 72 pages. I know it only through Professor H. von Mzik's review in OLZ., 1942, No. 5, p. 240. The author seems to believe in the bona fides of Abu Dulaf, but the reviewer calls his report mixtum compositum and concludes that we must see in it “ eine Falschung... von der es nur fraglich ist, ob wir sie Abu Dulaf selbst oder einem Spateren zuschreiben sollen ”.

⁴ Sub verbis: Turkistn, i, 840. lines 1–15: Nushajn, iv, 823

⁵ On this discovery see Validov, “ Meshedskaya rukopis Ibn ul-Fakiha ”, Izv. Ross. Akad. Nauk, 1924, 237–248, and A. Z. Validi Togan, “ Ibn Fadlan's Reiseberieht”, Leipzig, 1939 (edited after the same Mashhad MS.).

¹ De Goeje, BGA., v, p. xii: “ diligenter e libro Ibn Khurdadhbehi hausisse ”. Cf. al-Fihrist, 154, on Ibn al-Faqlh: ” Kitb al-buldn.nahwa 1000 waraqa: akhadhahu min kutub al-nas wa salakha kitab al-Jyhdni

² His copy of I. Faqlh was apparently more complete than that discovered in Mashhad, cf. para. 14 of our text as compared with Yqut, i, 841–3.

¹ Both of them came into their full rights only at the beginning of the tenth century (G. Haloun).

² Added in YqBt

³ Afterhas

⁴ MS

⁵ Yq.

⁶ Yq.l

⁷ Yq. omits.

⁸ MS. Yaq. omits.

¹ MS.

² Yaq. omits words after

³ MS.; Yaq.

⁴ Yaq.

⁵ Yaq. adds

⁶ Thus Yaq.; MS

⁷ MS.

⁸ Yaq.

⁹ MS.

¹⁰ Yaq.

¹¹ I. Khurd., 31, Yaq., iv, 823, is also correct, but i, 840, misunderstood:

¹² Paras. 6–12 omitted in Yaq

¹ Cf. Yaq., iv, 823.

² MS.perhaps

³ MS

⁴ MS

⁵ MS

⁶ Fide OTtpra 8?

¹ MS

² Apparently, cf. Qudama, 262

³

⁴ MS

⁵ Sic.

¹ Sic On the interchange of the terms madina and nhiya, cf. Ist., 225,, and I. Hauqal, 285 (but in the Kramers edition 396: al-nahiya al-madlna).

² Better basinatun “ fortified ”, as in Yqut

¹ Perhaps: *yajruna “ they fare ”.

² Maslib, but, more likely, this is but a vulgar spelling for maslih, “ military posts ”: on such confusion see de Goeje, BGA., iv, 282.

³ I. Kh., 28, K.wak.t, Qudama, 209, K.wyk.t; read both *Kuvekat.

⁴ Here end the quotations from Tamim.

¹ Para. 13 is full of confusion. The Toghuzghuz king invaded Transoxiana, and not China. The record possibly refers to the events of A.D. 821 (see below, p. 302), but Abul-Fadl must have mixed them up with the Chinese-Muslim fights in A.D. 751 when the Chinese artisans were captured, see my edition of Marvazi, p. 67.

² Para. 14 is also found in the abridgment of I.F., BOA., 32910, but in much more detail in Yaqut, i, 840–2, who refers to I.F. as his authority.

³ See my previous analysis of it in the Hudud al-'Alam, pp. 265–7, into which I now introduce considerable alterations.

¹ Kül-tegin, E 19.

² The Shine-usu inscription, see Ramstedt, “ Zwei Uigurische Bunneninschriften”, Journ. de la Soc. Finno-Ougrienne, vol. xxx, 191

³ A. N. Bernstam, Social and Economic Organization of the Orhhon-Y enisey Turks (in Bussian), Moscow, 1946, p. 53.

⁴ See Schlegel, “ Die chinesische Inschrift auf dem Uigurischen Denkmal in Kara-Balgasun ” Mim. de la Soc.Fin.-Ougr., ix, 1896. Cf. Chavannes-Pelliot, “Le traite manicheen ”, Journ-As., 1913, Janvier, 177–199; O. Hansen,“ Zur soghdischen Inschrift von Karabalgasun ”, Journ. de la Soc. Fin.-Ougr., xliv/3, 1930.

¹ F. Hirth, Nachwort zur Inschrift von Tonjukuk, St. Petersburg, 1899, p. 36; Chavannes, Documents, 34

² According to Maqrizi, Khital, i, 31335, Tulun, the father of the founder of the Tulunid dynasty in Egypt, was a Toghuzghuz. Among other tribute, he was sent to Ma'mun by the Samanid ruler of Bukhara, Nuh. b. Asad, in 200/815–16. More exactly this must have happened after 204/819 when Ma'mun gave Samarqand to Nuh.

³ MasüdI, i, 288 (on his confusion under i, 305, see below); 1st., 10 (the Toghuzghuz between Tibet, the Kharlukh, the Khirkhiz, and China); Gardizi, in Barthold's Otchot, pp. 90–2; Hudud al-Alam, † 12; Marvazi, p. 78.

¹ Possibly *Adhiish, I. Kh. 31.

² Perhaps *Majghari, the Bashqirs

³ One of the “ Turkish ” opponents of Bahrain Choben, see Nöldeke, Gesch. d. Perser, 269. In point of fact Shava (Shavagh) was a Hephthalite, Marquart, Eransahr, 83–4

⁴ Bernstam now places Suyab to the east of the Chu, but there may have been some confusion between two towns, for I. Kh., 28, and Qudama, 200, mention the “ town (or village) of the Turkish khaqan ” to the west (?) of the Chu river “.

⁵ The Sha-t'o were descended from the Ch'u-yiieh tribe which was incorporated by the West Turks, but seemingly did not belong to their inner federation. The original territory of the Ch'u-yüeh in the early seventh century was round Guchen, to the east of the tribe Ch'u-mi

¹ Through Abu Zayd, the author of the appendix to the so-called Sulayman-the-Merchant ? Mas'udi (i, 353–6) met him in Basra in 303/915.

² Still later, when the Uyghurs became known under their own name, the identity of the Toghuzghuz and Uyghurs was no more recognized and the Toghuzghuz of the past were treated as a vague, separate group, still extant, see Marvazi, p. 29.

³ WZKM., 1898, xii, 170. See also Komanen, 201

⁴ See also Hirth, F., Nachwort, pp. 34, 37 (“ vielleieht ”).Google Scholar

⁵ Articles “ Ghuz ” and “Toghuzghuz” in Enc. of Islam; 12 Vorlesungen, 54. Western Turks definitely had 10 clans (on-oq) and could not have been called Toquz-Oghuz

⁶ Streifzüge, 80: “ hier (am Orchon) kennt sie noch Ibn Churdadhbih, 30 “; but later, 390: ” Wie an anderer Stelle naher dargelegt werden soil, zeigt eine eingehende Analyse dieses Berichtes (i.e. of Tamim), von welchem I. Churdadhbih abhangig ist, dass unter der Hauptstadt der Toyuz oyuz hier unzweifelhaft bereits Kau-C'ang (beim heutigem Turfah) zu verstehen ist.”

¹ Otchot, 34; 12 Vorlesungen, 54; El., s.v. Toghuzghuz

² See Mujmal al-tatvanlfk, ed. Bahar, 1318/1939, p. 99, on the Issik-kul as the original home of Turk's family. Cf. Hudud al-'Alam, p. 292.

¹ A. N. Bernstam, “ Archaeological Survey of Northern Kirghizia ” (in Russian), Frunze, 1941, pp. 80–2.

² Hiouen-Thsang, Memoires sur les contrees occideiUales, tr. par S. Julien, ii, 11–12

³ Cyprinus carpia, Squalins Schmidtii, Schizothorax argentatus, but neither Hsuan-tsang's traditional “ dragons ” nor even Tamim's “ sea-animals ”.

¹ Now ruins of Talas, circa 280 km. to the north-east of Tashkent.

² Bernstam places Navekat near the village Orlovka, to the west of the Chu. But cf. Hudud, 290–1, 298.

³ See Hudud al-'llam, p. 273 († 12, 3) and p. 293 (§ 15, 13).

⁴ Here we would expect some such words as: “are the Toghuzghuz. And the frontier-point of the Turks] is called (Lower) Nushajan.”

¹ The latter or his scribe, thinking of the situation in the tenth century, has mis-spelt naving in mind *Boghra-khan, the ancestor of the Qarakhanids, whose home was Kashghar, see Hicdud, 278, but even so the distance indicated is too short.

² See Hudud, p. 273 († 12, 3) and p. 293 (§ 15, 13). Qudama has many points in common with Gardlzl.

³ Reproduced verbatim in I. Faqih, BGA., 3287, and in Yaqiit, iv, 823 (under the heading Nushajan).

⁴ See Baladhuri, ed. de Goeje, Vocabulary, 52; Dozy, Supplement, i, 666; Lane, p. 1387. In analysing Arab itineraries, east of Taraz, Bernstam comes to the conclusion that 1 farsakh = 6 km.

¹ According to the Chinese sources the distance from Khara-balghasun via Chung shouchiang ch'eng (on the Yellow River, in the Ordos bend) to Ch'ang-an (the capital) was under 4,000 li = 2,450 km. (Haloun).

² The lake which the Chinese envoy Wang Yen-te (see Journ. As., 1847, ix, 62) mentions near Bish-baliq, was apparently in the nature of a pond serving for the purposes of amusement. Cf. Hudud, 269

³ “ Capable of holding (tosa') 100 men.” Thus in the Mashhad MS., in I. Kh., 31, and in Yaqflt, iv, 823. In Yaqut, i, 840, the statement has been wrongly read: “ The king's tent is of gold and on the top of his castle (there are) nine (tis') hundred men.” The passage on the tent has been also incorporated in Abii-Dulaf's spurious report (see Qazwini's quotation from Abii- Dulaf, Athar al-bilad, 391: “ the tent contains 1,000 men ”), and the additional passage from an aberrant version of Qazwini (Gotha, Cod. No. 231) which Wustenfeld translated in Zeilschrift fttr vergleichende Erdlcunde, ii, 1842, p. 212: “ (the khaqan) has a golden tent on the summit of his castle; it contains 100 men and can be seen from a distance of 5 farsakhs. Their banners are black. We sic] travelled among them twenty days in great fear.” Even the number of days may be an echo of Tamlm

¹ See Chavannes, Documents, 305

² I found it in the Russian translation of Hyacinth, i/2, 449–450, but my friend Professor G. Haloun has kindly retranslated for me the whole passage, ch. 217/ii, fol. 20a, which I have summarized.

³ Repeated in I. Khurdadhbih, 31; Qudama, 262; and Yaqut twice: i, 840, and iv, 843.

⁴ Radloff, Atlas der Altertumer der Mongolei, pi. xxvii.

⁵ According to Dolbezhev, “ V poiskakh Bish-balika ”, ZVO., xxiii, 1915, p. 14, the town had three gates (?); Sir A. Stein, “Innermost Asia”, iii, plates 23 and 24 (nothing conclusive on this point).

¹ The Shine-usu inscription (W 5) refers to the town of Bay-baliq which the qaghan let the Soghdians and the Chinese build on the Selenga.

² I am grateful to Professor Haloun for a reminder that in Inner Mongolia, in former Onggut territory, there are traces of considerable medieval sites of agriculture (O. Lattimore, D. Martin), and the same applies to the ancient Qirghiz area (on the Yenisey). Unfortunately we know nothing on the exact route followed by Tamim.

³ According both to the T'ang-sku (Hyacinth, i/2,446) and to GardizI, in Barthold, Otchot, 87 (transl. Ill), the Qirghiz burnt their dead, which in the eyes of the Muslims was a symbol of fire-worshipping. Thus the Rus (ancient Norsemen) were also called majus (“ Magians ”), see Minorsky, “ Rus ”, in El.

⁴ See Chavannes-Pelliot, “ Un traite Manicheen ”, Jau.rn.As., 1913/1, pp. 261–303. Cf. below the story of al-Fihrist, 377

⁵ Barthold, Otchot, 90–2 transl. 114–16

¹ The Hudud was written in A.D. 982; the source of the passage is probably Jayhani, who wrote in the earlier part of the tenth century

² For another indication of the Qarluq-Uyghur struggles see Hudud, † 15, 12.

³ See Marquart, Streifz's mind. He had in view the eclipse of the Uyghur power on the Orkhon (A.D. 840), and he seems to quote the Barskhan-Qarluq struggles to which Tamim referred en passant while travelling to the Toghuzghuz capital perhaps circa 821.

⁴ After 809 settled in Northern Ordos, see above, p. 288, n. 5,

¹ In the first case by the Qirghiz (still on the Yenisey); in the second by the Qarluq and the remnants of the Tiirgish federation.

² More exactly from the villages *Kuvekat, see 12.

³ Hirth, “ Nachwort”, 36.

⁴ A similar result might be obtained by adding up the original fifteen tribes of the Tolis federation (one of which were the Uyghurs themselves, Hyacinth, i/2, 426–442; Chavannes, 87–8), with the Qarluq and Basmil.

¹ Hyacinth, i/2, 382–426; Chavannes-Pelliot, op. cit., 286

² Hyacinth, i/2, 403, 418; Chavannes-Pelliot, op. cit., 277. According to Wang Yen-t6, who visited the Uyghurs of Turfan in 981–3, Journ. As., 1847, ix, 64, the price of a good horse was only one piece of silk; that of a common horse 1 ch'ang (3 metres of silk).

³ In any case, in 827 Princess T'ai-ho, a very young widow, was still in the horde, see below, p. 303.

⁴ Bretschneider, Medixval Researches, i, 242.

¹ Even at a low rate of 30 km. a day, the traveller would have covered 2,700 km. in ninety days.

² It was already known that Schlegel wrongly took the “ heavenly qaghan ” for the Emperor of China see Chavannes-Pelliot, “ Le tratte manicheen ”, JA., 1913, p. 198.

¹ Chinese face, lines 17–21; ef. Soghdian face, lines 19–20 (fragments only).

² My friend, Dr. G. Roerich (letter 2. vii. 1947) suggests that J.h.run may represent the first half of the name of the Tibetan king Khri-sron Ide-btsan (755–797), the initial kh having a palatal pronunciation in some Tibetan dialects (Amdo).

³ See above, p. 288: Mas'udI, i, 288.

¹ Barthold, Turkestan, 176.

² To this element belonged the famous founder of the Saffarid dynasty Ya'qub ibn Layth.

³ Tabari, iii, 631 (year A.H. 178): 20,000 of the 'Abbasiya were sent to Baghdad where they were surnamed al-Karanbiya “ cabbage-(eaters) ”; the others were left in Khorasan on their own registers. The verse of Marwan ibn Abl-Hafsa suggests that the numbers of the original corps were somewhat exaggerated.

⁴ Incidentally, the very mutilated para. 13 in the Mashhad MS. may have some relation to this very series of events in Ushriisana, although the “ Toghuzghuz ” incident seems to be different from the incident of “ Fadl and the Turks ” described in Tabari

¹ Marquart, WZKM., 1898, 163; Barthold, 12 Vorlesungen, 55.

² The name Tamim occurs in the family of Nasr ibn Sayyar, whose grandson Rafi′ ibn Layth raised a revolt in 806–810 and sought help from various Turkish tribes. Even before 775, Layth (ibn Nasr ?) was charged with a mission to the ruler of Farghana who lived in Kashghar.

³ On them see my commentary on the Hudiid, pp. 304–310.

¹ See above, p. 287, the quotation from Rashld al-DIn.

² De Goeje, BGA., vii, p. viii. [This conclusion, valid for the time of our documents, leaves aside the question of the possible alternance r/z, which is a characteristic feature of Turkish phonetics. Originally, and tribally, Oyaip, Oghuz, and perhaps (?) Uyghur may be variations of the same name.]