^{page 181 note 1} Massé, H., “Persian inscriptions,” in Survey of Persian Art, ii, 1794–1804Google Scholar, deals chiefly with later periods.

^{page 181 note 2} Cf. de Bode, C., Travels in Luristan, 1845, ii, 251Google Scholar, with a plate.

^{page 181 note 3} Transl. by Sir D. Ross, p. 416.

^{page 182 note 1} Cf. also Ettinghausen, , “Dated faience,” in Survey of Persian Art, ii, 1667–1696Google Scholar.

^{page 183 note 1} The earliest of the Chinese documents found at Dandan-Uyliq is of a.d. 758, see SirStein, A., JRAS., 1903, p. 745Google Scholar.

^{page 184 note 1} See Hoernle, , “A report on the British collection of Antiquities from Central Asia,” J.A.S. Bengal, lxx, part i, extra No. 1, pp. 26–8 (1898) with a good photographGoogle Scholar.

^{page 185 note 1} Margoliouth: “Husain, son of Liko-Kongo, my maternal uncle, in the year 40 (of his age).”

^{page 185 note 2} Margoliouth: “Almayah (?), a village in Nikatanj (?), part of the boundary of the land of. . . .”

^{page 187 note 1} Ibn al-Athīr, xi, 55, cf. Barthold, , Turkestan, 322Google Scholar.

^{page 187 note 2} We can imagine some name like *Lin-huo-k'wei . The Chinese scholars whom I have consulted admitted that the name sounds Chinese but abstained from making any definite suggestions.

^{page 188 note 1} We can hardly suspect in them any trace of Christian (Neetorian) influence. No importance either can be attached to the crosses used by the illiterate witnesses instead of signatures.

^{page 188 note 2} With a parasitic 'ayn, cf. la'l for lāl, ka'k for kāk. C. F. Andreas explained the Arabic name of Erzerum Qālyqalā as a compound of. the Armenian name of this place Karin (Karnoy, Karnay) + kalā, see Hartmann, M., Bohtān, 1897, p. 145Google Scholar.

^{page 190 note 1} Or L.ngūkūhī in accordance with Persian phonetics.

^{page 192 note 1} These details are borrowed from SirRoss's, D. postscript to Barthold's article “The Bughra khan mentioned in Qudatqu (read: Qutadghu) bilik”, BSOS., iii/l, 1923, pp. 151–8Google Scholar.

^{page 192 note 2} Prepared under the auspices of the learned qādī of Yārkand, native of the great Islamic centre of Bukhara.

^{page 192 note 3} Sir D. Ross read mubaddhira and translated “producing 30 ass-loads of wheat”, but this is contrary to the parallel texts in Turkish and to the estimation of land according to the amount of seed still practised in Central Asia.

^{page 192 note 4} The copy is in a good Muslim hand but the copyist apparently did not understand Turkish. Under the text there is a Persian translation by Sir D. Ross's munshi who knew Eastern Turkish but did not know the meaning of older terms. I have also the original decipherment by Sir D. Ross, but it only partly covers the text.

^{page 193 note 1} The word bwrtgha (?) standing before nāk cannot be right. I am inclined to read türk-jä “in Turkish”, in view of the word bil-turkiyya which introduces the Turkish yond-yïlï in an Arabic document of the same collection.