^{page 221 note 1} See, for example, the introduction to the collection of administrative documents of al-Juvaynī, Mu'ayyad al-daula Muntakhab al-dīn Badī' Atābak, in ‘Atdbat al-kataba, recently published by Qazvini, M. and Eghbal, Abbas, Tehran, 1329/1950, pp. 1–5Google Scholar. The said Mu'ayyad al-daula was the maternal uncle of Juvaynī's great-grandfather Bahā al-dīn.

^{page 222 note 1} Juvaynī briefly alludes to these events in the previous chapter (I, 222) in which he says that, after the advent of Ögedey, Batu subjugated the neighbouring territories consisting “of all that remained of Qipchaq, the Alān, the Ās, and the Rūs, and other lands, such as Bulghār and MKS”. This reference is definitely vague and approximate (V. M.).

^{page 222 note 2} It is at this place that I assume a great lacuna in the text provoked either by the desire to abridge the report or by a μοιοτλευτον (V. M.).

^{page 223 note 1} Read: “nothing but its namesakes” (V. M.).

^{page 223 note 2} It is “the horsemen (who marched against) the Kilar and the Bashghird” (V. M.).

^{page 223 note 3} In 1246 Pian de Carpine saw in Bati's camp on the lower Volga “tents made of linen. They are large and-quite handsome, and used to belong to the king of Hungary”. See Rockhill, 's trans, in Hakluyt Society, 2nd series, no. iv, p. 10.Google Scholar

^{page 223 note 4} Cf. a very short enumeration of the “eleven nations” of the West in the so-called “Secret History of the Mongols”, §§ 262, 270, 274 (I have used also the recent translations by Kozin, S. A., Haenisch, E., 1941Google Scholar, and Grønbech, K., 1945)Google Scholar. The Far-Eastern texts, both Mongolian and Chinese, are very brief.

^{page 224 note 1} The war prevented Pelliot from consulting this publication.

^{page 225 note 1} *Kiral, which suits better Hung, , királyGoogle Scholar “king” than Polish król. For the metathesis see the name of the river in Mongolia, Kelüren/Kerülen.Google Scholar

^{page 225 note 2} This paragraph is not in Juvaynī.

^{page 225 note 3} *Ba-jerge “in an encircling movement, like at a battue”. Juvaynī, III, 10Google Scholar, uses another form of the word: nerge. In Turkish n (in some dialects) corresponds to y (j).

^{page 225 note 4} This episode corresponds to Juvaynī, III, 10–11Google Scholar, who omits the names Ūlīrlīk and Qāchīr-Ūkūla.

^{page 225 note 5} I.e. the Ossets, but the name of the prince has not been identified. In Mongol khachir means “a mule”, but a Mongol name among the Ossets is unlikely.

^{page 225 note 6} Here the editor supplies six lines of missing text from Juvaynī, see Qazvīnī's, ed., iii, 10–11.Google Scholar

^{page 226 note 1} This record is not in Juvaynī. It opens the account of the campaign in the northern zone.

^{page 226 note 2} The Muslim year 634 corresponds to 4th September, 1236–23rd August, 1237, but the Mongolian year (beginning in January-February) began in the later part of the Muslim year. Practically it corresponded to the Christian year 1237, as confirmed by the chronology of the events in Russian annals.

^{page 226 note 3} Only here, after a great gap in Juvaynī's text, do we catch up with the final part of his report.

^{page 226 note 4} Juvaynī's MKS.

^{page 226 note 5} Not in Juvaynī.

^{page 226 note 6} In Mongolian, “Iron Gate”Google Scholar, i.e. Darband.

^{page 227 note 1} This date is wrong. Further down Rashīd himself seems to admit that during the operations of 1241 the princes had not yet heard of the death of the Qā'ān. According to Juvaynī, I, 159Google Scholar, Ögedey died on 5 Jumādā, II, 639/11th 12, 1241.Google Scholar

^{page 227 note 2} This is a second and much more detailed version of what Rashīd had copied from Juvaynī (see above).

^{page 227 note 3} According to Pelliot, , La Horde d'Or, p. 153: Moldavians.Google Scholar

^{page 227 note 4} The Chinese Yüan-shi also speaks of five roads followed by the Mongols, see Bretschneider, , I, 331Google Scholar

^{page 227 note 5} Different from Qūqdāy mentioned under a.d. 1240.

^{page 232 note 1} The same remark applies to some other passages. See the exposition of the Ismā'īlī doctrines in Juvaynī, iii, and in Rashīd al-dīn, quoted by Levy, R., JRAS., 1930, 509–536.Google Scholar

^{page 232 note 2} Recently: Minorsky, , Ḥudūd, 1937, p. 446Google Scholar: “M.k. mentioned [in Juvaynī, 222] together with Bulghār seems to refer to the Moksha (a Mordvan tribe)”; Pelliot, , La Horde d'Or, 124Google Scholar, distinguishes between the two towns bearing similar names but finally takes Juvaynī's *Makas for Moscow (in some contradiction with his former and correct statement in Jour. As., Avril, 1920, pp. 168–9).Google Scholar

^{page 232 note 3} The text can be read only as “juz ham-nām-i ān naguzāshtand”, ham-nām being “namesake”.

^{page 232 note 4} Yüan-shi, 2, 7a, and 122, 13bGoogle Scholar: Miä-khiä-sï; 128, 14bGoogle Scholar: Mai-khiä-sï; 132, 9aGoogle Scholar: Mai-ko-sï. The variants of the first syllable also occur in the transcription of Ma- in Märkit, see Pelliot-Hambis, , Histoire des campagnes de Gengis khan, 1951, i, 217Google Scholar. (This note is by my lamented friend, Professor G. Haloun (d. 23rd December, 1951), who was ever ready to help me with his advice on Far Eastern matters.)

^{page 232 note 5} Bretschneider, , Mediaeval Researches, i, 309–317.Google Scholar

^{page 232 note 6} In his translation Haenisch, E., p. 188Google Scholar, wrongly explains Meget, Meket = Mekes, as “die Hauptstadt Georgiens, Mzchet bei Tiflis”. In 1239 the Mongols did not cross the Caucasus and Mtskheta has nothing to do with the Ās.

^{page 233 note 1} Both Alān and Ās refer to the same people, the ancestors of the present-day Ossets (Ous-et'i in Georgian, “the country of the Ous,” i.e. Ās). Alān seems to be the north-Iranian form of “Aryan” (ry > l). The reason of the double appellation Alān/Ās is not clear. It is possibly due to the existence of two cognate tribes which formed the Osset people, which even now speaks three different dialects.

^{page 233 note 2} And most probably over some neighbouring tribes of the Northern Caucasus, such as the Chechens.

^{page 233 note 3} Cf. Rusta, I., p. 147Google Scholar, and Gardīzī, in Barthold, , Otchot, p. 101.Google Scholar

^{page 233 note 4} Mas'ūdī's statement to this effect is confirmed by such names of the rulers as Bukht-Yishō' found in the old history of Bāb al-abwāb (compiled before 500/1106) which I am publishing.

^{page 234 note 1} Les peuples du Caucase, 1828, p. 23.Google Scholar

^{page 234 note 2} Urukh is a left tributary of the Terek. On the Urukh too lies a place called Moska (apparently insignificant).

^{page 234 note 3} Gan, K. D. (Hahn), Sborn. opisaniya mest. i piemen KavkazaGoogle Scholar, quoted in Uvarov, Countess's Materialī archeologii Kavkaza, 1900, viii, 254Google Scholar. On the other hand, according to Pfaff (see Miller, , l.c., p. 36)Google Scholar, the capital of the ancient Alān kingdom should be looked for on the Fiag-don, which flows between the Urukh and the Terek.

^{page 236 note 1} In the Troitsk Chronicle, as restored by Priselkov, (published in 1950), p. 356Google Scholar: “Tsar Ozbek” killed Michael “on the river” Nay a, near the town Dyedyakov. Here ΗΑИ can be only a bad reading of the last letters of CebeΗЦВ.

^{page 236 note 2} The “Iron Gate” is a classical name for Darband, but (a) here it may be used only to indicate the approximate direction, and (b) we have instances in the Russian Chronicles when the Iron Gate refers indifferently to the Caucasian chain. In the life of Daniel of Galicia it is said that “he chased khan Otrok (Atraq ?) into the Obez land (Abkhazia) beyond the Iron Gate”. Abkhazia lies at the westernmost end of the Caucasus, some 600 kilometres from Darband, as the crow flies.

^{page 236 note 3} The “statue, image” (though said to be of copper) is likely to be a Turkish balbal (from which Russian bolvan and baba (kamehhaЯ БaБa) are derived), i.e. a stone image of an enemy placed at the funerary mound of a Turkish hero. Niẓāmī (who died in 605/1209) in his Iskandar-nāma (composed in 591/1200), Tehran, , 1316/1937, 427–8Google Scholar, says that when Alexander marching against the Russians penetrated into the steppe of Qifchaq, he was shocked by the freedom of the Qipchaq women going about unveiled. His sage made out of black stone a talisman in the shape of a veiled maiden, and the women took example from her. “That talisman still stands there”: the Qifchaqs approaching it bow before it; a horseman deposits an arrow in its honour (dar kīsh-i ū) and a herdsman offers a sheep to it which is devoured by the eagles hovering over it. Some vague memory of the image may have survived even in the Tractates de duabus Sarmatiis (1517) of the learned Polish doctor Maciej z Miechova, who (Part II, ch. iv) states that beyond Viatka [sic], in Scythia, there exists a great idol called Zlotababa (“Golden woman”) worshipped by local tribes who make offerings to it, be it of a hide or even of a hair, after which the visitor “inclinando se cum reverentia pertransit”. This latter report may have been influenced, however, by the report of the Russian chronicle of Khlīnov on the capture by Novgorodians (in 1174) of a town on the Viatka, called Bolvansky, because of a heathen statue (БοΛΒaΗ) found in it.

^{page 236 note 4} The identity of this river is not clear. According to Karamzin, , ed. 1842, ivGoogle Scholar, note 237, this is “the Gorkaya which flows into the Caspian”. On the recent maps a Gorkaya (“Bitter”) river belongs to the basin of the Manīch, to the north-west of the Kuma basin. This seems to be the *Ajī mentioned in the Chronicle.

^{page 236 note 5} Ed. Defrémery, , ii, 375Google Scholar. It is tempting to identify Üzbek's wife Bayalshi (?), who according to the Chronicle saved Michael's followers, with Özbek's wife Bayalūn, a Byzantine princess whom Ibn-Baṭṭuṭa accompanied to Constantinople where she was going for her confinement. This journey is supposed to have taken place towards 1334 (?), and in this case Bayalun must have been very young in 1318. [Correction. Pelliot, , l.c., 84–5Google Scholar, avers that the name Bayalun was borne by Üzbek's mother, and possibly by two of his wives, of whom the first died in 1323.]

^{page 237 note 1} Which at the latitude of Vladikavkaz flows nearly parallel to the Terek, but then swings to the N.E. and only past Grozny joins this major stream.

^{page 237 note 2} Karamzin, , iv, note 157 (p. 59)Google Scholar, says that Dyedyakov probably corresponds to the “Diven, or Dedukh”. These names are not on the present-day maps. A Datikh is shown in the Chechen country on the Fortanga flowing to the east of the Assa, see Baddeley, J., The rugged flanks of Caucasus, 1940, ii, map v.Google Scholar

^{page 237 note 3} On him see now the Georgian novel by Dadiani, Shalva, Yuri Bogolubsky (Buss, , trans., Tbilisi, 1951).Google Scholar