^{page 641 note 1} Perhaps rather ‘ untrodden ’.

^{page 641 note 2} See also Note on p. 644.

^{page 641 note 3} Wackemagel, Ai.Or., ii, 1, p. 110.

^{page 641 note 4} Walde-Pokorny, Vergl. Wb. der idg. 8pr., ii, p. 24, suggest that these last two are to be related more directly with the verbal root seen in Skt. pádycte, etc

^{page 641 note 5} Buck, Or. of Oscan and Umbrian², p. 62.

^{page 641 note 6} Wackemagel, Ai.Or., iii, p. 235. But these are in fact of later occurrence than the corresponding forms with -pad-.

^{page 641 note 7} Wackemagel, op. cit., p. 236. H. Frisk, Zur indoimnischen u. griechischen Nominalbildung, p. 53, can scarcely be right in proposing upabd- as a possible secondary formation from thematic *upabda-. But the v.l. upabda might be conceived of as loc. sg. of an -i- stem (cf. Skt. upabdi)

^{page 642 note 1} Duchesne-Guillemin, Les Composts de VAvesta, p. 181.

^{page 642 note 2} Professor H. W. Bailey, who has drawn my attention to this, remarks that just as abda occurs in a passage connected with snow, so Oss. sevsed as an epithet of snow means ‘ without footprints, smooth’.

^{page 642 note 3} Skt. ápada- n. ‘ where one should not come’ or ‘ that which is no place for’ (AV. vi. 29.2) is a compound of padá- n. ‘ place ’.

^{page 642 note 4} Communication from Professor H. W. Bailey

^{page 642 note 5} Op. cit., p. 53.

^{page 642 note 6} Op. cit., p. 35, but on p. 143 he posits the athematic form

^{page 643 note 1} The queried feminine papadá of PTS. Dictionary appears to be due to the same mistake as the lexicographers made for Skt. prápad-.

^{page 643 note 2} Wrongly connected with Skt. pádma- in my Nepali Dictionary, 389 6 15.

^{page 643 note 3} From a MS. vocabulary of Khowar collected by Lieut.-Col. D. L. E. Lorimer

^{page 644 note 1} The only other instance I know of a word with Skt. -dm- surviving in modern Indo-Aryan is VS. ddnum- n. ‘ flood ’, Pān. odma- m. ‘wetting ’, with -bb- in Kumaoni wob ‘ wetness ’, and -dd- in Pkt. odda- ‘ wet’, Hi. od m. ‘wetness’, odd ‘wet’, Bg. odd ‘wet’, Ass. odd ‘half-dried’, Or. oda ‘wetness’, odd ‘wet’. This fourth development, -dm- > -dd-, appears in Skt. chddmann.‘ roof, covering ’: Pa. chadda-n. ‘roof, veil’ (Pkt. chailma-, choma-, chamma-m.n.‘ covering, deceit’); Bg. chad ‘roof’ (S. K. Chatterji, Bengali Language, p. 506, wrongly <; chattra-); *avacchadman-: Mar. osad -t f. ‘shade’. If, as evidenced above, -dd- represents the regular development of -dm- in Hindi and Bengali, these languages may have adopted the forms based on paduma- (< pddma-) to avoid a clash with the descendant of parda-.