^{page 889 note 1} jīva may represent Skt. jīva or jeta.

^{page 890 note 1} Line 32 = 34 crossed out.

^{page 890 note 2} drya = drū.

^{page 890 note 3} Blurred akṣara (? na).

^{page 892 note 1} Blurred, probably part of salī.

^{page 892 note 2} Blurred, uncertain.

^{page 892 note 3} mya = mu.

^{page 892 note 4} Blurred pū.

^{page 892 note 5} Blurred.

^{page 892 note 6} Read ru ?

^{page 893 note 1} ṣte put for ṣṭāre.

^{page 893 note 2} Or Banāca (-vā being then loc. plur.).

^{page 894 note 1} Uncertain akṣara: ṣ is clear and -i.

^{page 894 note 2} Opaque paper pasted over makes parts illegible, but the words can be supplied from the repetitions.

^{page 894 note 3} Subscript.

^{page 894 note 4} Blurred out.

^{page 894 note 5} hūrī written over bāyī.

^{page 894 note 6} Subscript.

^{page 896 note 1} In the Cambridge University Library, Kanjur only. For translation see Thomas, T. i 19–20.

^2–7 Narthang variants:—

^{page 896 note 2} mi-ḥpham-pa.

^{page 896 note 3} khyim-ćhig.

^{page 896 note 4} dbyiṅs.

^{page 896 note 5} ḥphreṅ.

^{page 896 note 6} yoṅsu.

^{page 896 note 7} gnasu.

^{page 896 note 8} Loc. cit. p. 97 Aparājita renders the Tibetan gźan-gyis mi-thub-pa ‘not conquered by others’. Note that in Professor Thomas's transliteration ẖ = Tib. ḥ.

^{page 896 note 9} gha-gha-na-sva-ra.

^{page 896 note 10} su-gar-na-ma-la.

^{page 896 note 11} leags-kyu.

^{page 896 note 12} sta-na-ba-ti.

^{page 897 note 1} ẖ = Tib. ḥ.

^{page 897 note 2} 'A-rya-ba-lo.

^{page 897 note 3} Ma-ni-bha-ba.

^{page 897 note 4} Translated by Sylvain Lévi, BEFEO 5, 268.

^{page 897 note 5} Top of akṣara lost.

^{page 909 note 1} gyastaṣṣai ‘devaputra’ is found also Ch. c. 001, 991 cạndrraprabhú gyastaṣṣai, 981 mahiśvarú gyastaṣai. Similarly Viśvakarman, Jap. bishukatsuma is both deva and devaputra, Hōbōgirin, s.v.

^{page 909 note 2} Pelliot, T'oung Pao 1933, 94 ff. treated of suei < z^wiḙ, to represent the first syllables of vairambha, vaiśālī, viśvabhu, (Pali) verañjā, viśākhā, and the bi of śibi and ve of uruvelā, and conjectured a fricative pronunciation of v, such as zv or žu.

^{page 909 note 3} Cf. also K 253, 235 iu-ien < *yužan, Skt. yojana, beside K 253, 845 iu-sü n < Jap. yu-en, yujun (Oda, p. 1757, 1760, and Pelliot, T'oung Pao 1933, 94 ff.).

^{page 909 note 4} As an epithet of Avalokiteśvara, śāntendriya occurs in the Suprabhāta-stotra 24 (ed. Minayev, Zapiski vostočnago otd'denia 1887).

^{page 910 note 1} Correctly here, but the syllables are wrongly divided in the later edition in the Journal of the Taisho University 1930 vi-vii, Die nordarischen Abschnitte der Adhyardhaśatikā-prajñāpāramitā, p. 60Google Scholar.

^{page 911 note 1} Kharoṣthī Inscriptions, ed. Boyer, Rapson, Senart, Noble.

^{page 911 note 2} The numbers refer to the pages of Thomas, T. i.

^{page 913 note 1} The form printed in BSOS ix 69 is unattested, being due to an ambiguous apparatus critious.

^{page 914 note 1} For ṣə < źia which is not in Karlgren I am indebted to Professor Haloun The fan-ts‘ie is

^{page 916 note 1} On ṣau, see BSOS x 599–600.

^{page 916 note 2} I should not have published the etymology quoted from me by Konow, Norsk-Tids. Sprogr. xi 55.

^{page 917 note 1} Ed. Oldenburg, , Zapiski vostočnago otd'elenia 1897-8Google Scholar, studied by Sylvain Lévi in connection with the Chinese and Tibetan versions, Journ. Asiat, 1915, Le Catalogue géographique des Yakṣa dans la Mahāmāyūrī. One name in this text, verse 72 vaikrtika the yakṣa of Gandhāra was compared by Lévi with the Av. vaēkərəta, but has been overlooked by recent writers on Zoroastrian problems. For -postā, read probably -posto?

^{page 918 note 1} Fan-ts‘ie = p‘i < p‘iei.

^{page 919 note 1} Cf. in‘the Milinda-pañha, Chinese K 342, 225, 1265 A kia-ie-uei < ka-úp-_wi *Kāśyapīya', K 484, 186 kü -i < ku-i ‘Kuśi’, see Pelliot, Journ. Asiat. 1914, Les noms propres du Milinda-paṅha, Demiéville, Les versions chinoises du Milindapanha. Note also K 235 ien < ún for žan = Skt. -jana, -cana, supra B I (6).

^{page 919 note 2} Yet a second correction is needed in the passage quoted there in note 1: read baudasattū rāṇdú ‘bodhisattva and king’.

^{page 920 note 1} Études asiatiques publiées à I'occasion du 25^e anniversaire de IWécole française de I'extrême orient ii 43.

^{page 921 note 1} Lüders, , Bruchstücke der Kalpanāmaṇḍitikā des Kumāralāta (1926) p. 34, note 7Google Scholar

^{page 921 note 2} This name and no-geḥdra-śil would be in Khotanese *nāgevdrasīha and *nāgendraśīla. They are in a document published in Hoernle, Manuscript Remains, p. 403. For sīha = Pkt. sīha, Skt. siṇha, cf. also the names P 4099,. 432 devedrraśūrasīha ‘Devendra-śūra siṇha and P 2027, 11 nāgaidrrārmúsīha'; and for nāgendra, cf. Or 8212 (162) *58 nāgaidratqrrda ‘Nāgendravardhana’. ‘The form sīuhú occurs in Kha 1, 56, 1 B 3, and as a Buddha name sīhau Bhadrak. 223.

^{page 922 note 1} aysmu = vijñūna here.

^{page 922 note 2} ūśrrīvye, inf. to a verbal base ūśrrav- ‘erect’, from Pkt. *uśrav- = ussav-,ūsav-Skt. ucchrāpaya-, causal to ud with śtay-.

^{page 923 note 1} Is this the place contained in the title sakarāja ‘King of Saka’ (Lüders, Zur Geschichte und Geographic Ostturkestans, p. 255,1. 7) ? The name of Yarkand in Chinese (Herrman, Atlas of China) K 847, 1187 so-kü < suâ-k ^wo, Ptolemy ooîra, var. lect. σoγα, oαγα bears considerable resemblance.

^{page 923 note 2} In English ‘bell-metal’ is an alloy of about four parts of copper to one of tin.

^{page 924 note 1} More details are given by Professor Thomas in Asia Major 2, 255 ff.

^{page 924 note 2} For -y-, -t-, -v- see BSOS x 572.

^{page 924 note 3} For the forms of kinnara in Khotanese, see BSOS x 583.