^{page 77 note 1} It is not excluded that in the Sanskrit virāma may have been at times intended by the sign transliterated i, even with tt.

^{page 83 note 1} Since the above paper was written the Rev. Rahul Sariikrityayana has announced in JBORS., xxi (1935), pp. 31 and 33Google Scholar, the discovery in Tibet of two incomplete MSS. of the Sanskrit text of the Vttaratantra. It is to be hoped that photographs will be made available for the publication of this important material.

^{page 84 note 1} samrritta written over

^{page 84 note 2} Struck out.

^{page 84 note 3} With -i struck out.

^{page 84 note 4} muhta=ma below the line.

^{page 84 note 5} With -ā struck out.

^{page 84 note 6} Below the line.

^{page 84 note 7} Uncertain, possibly ra.

^{page 84 note 8} With -au struck out.

^{page 85 note 1} °Śalyaṁ would be preferable metrically.

^{page 85 note 2} Or trātvā.

^{page 85 note 3} The restoration is doubtful.

^{page 85 note 4} The roll reads wrongly duḥkhitasyo°.

^{page 85 note 5} Did the writer intend anāś;itam ?

^{page 85 note 6} It looks as if anagrimam is meant here.

^{page 85 note 7} Is deśayan mahāsaṁbuddhas intended ?

^{page 85 note 8} tha-ma, T., i.e. ante.

^{page 86 note 9} bā; with ā probably struck out.

^{page 86 note 10} Struck out.

^{page 86 note 11} ryü written under rū (with ū struck out).

^{page 86 note 12} tta struck out, ca written below.

^{page 86 note 13} sti … ca struck out, = 1. 16.

^{page 86 note 14} ri with -ī written above.

^{page 86 note 15} tti … vatti below the line.

^{page 86 note 16} bū struck out, bhū beneath.

^{page 86 note 17} An uncertain aksara struck out, ye written below.

^{page 86 note 18} rvi with -i struck out.

^{page 86 note 19} prasimrritte struck out.

^{page 86 note 20} ysam ntha uncertain.

^{page 86 note 21} ttrām struck out, with ttām below.

^{page 87 note 9} T has , possibly for svatāsaṁpatti°.

^{page 87 note 10} ma-rigs sgribs-pa, T, i.e. avidyāvaranesu, which is the better reading.

^{page 87 note 11} The Chinese and T transpose b and c, rightly as the continuation shows. The three following characters do not belong to the verse and may represent a misplaced uktam.

^{page 87 note 12} T has ldan-pa “possession”; the only possible word, prāppti, is metrically difficult. For pattau from pad, which is free from this objection, there is only the authority of the Indian lexicons.

^{page 88 note 22} krri with -i struck out.

^{page 88 note 23} ri struck out.

^{page 88 note 24} pū … ttau below t h e line.

^{page 88 note 25} Uncertain.

^{page 88 note 26} ttatti below the line; karūṃ … mayam struck out.

^{page 88 note 27} rya first written, later stroke added to left.

^{page 88 note 28} ptī below the line in thinner handwriting.

^{page 88 note 29} vi with -i struck out.

^{page 88 note 30} ra struck out.

^{page 88 note 31} Uncertain akṣara struck out, with ṇe below.

^{page 88 note 32} rahna to end of line struck out.

^{page 88 note 33} rāṃ with ryāṃ, below.

^{page 88 note 31} uncertain sign.

^{page 89 note 13} , T, i.e. saṁkleśavyavadāna°. It omits bhūmiṡu and jagat, and reads °pūrvaparã°. The Chinese had bhūmi and paraphrases the preceding words “all kinds of tendencies (sui, Giles 10396, ? anuśaya) and abandonments ”, reading perhaps citrā° for cakrā°.

^{page 89 note 14} prakarana from kī, certified by the Chinese's san “disperse”; one would expect , possibly indicated by T's gzhig.

^{page 89 note 15} legs-gnas, T. i.e. sustho; so also the Chinese.