¹ § 12 fin.

² § 2143.

³ AiW s. v.

⁴ GIP 1. § 382 fin.

⁵ Ibid. §413.

⁶ Ibid. § 414. For mazdā as possibly instr. in form, cf. AF 3.55f.; Easton JAOS 15.202.

⁷ § 46.

⁸ I.e., 12 in all, cf. table accompanying § 51.

⁹ §41.1

¹⁰ Cf. § 21 init.

¹¹ Gdr. 2.2. § 112.4).

¹² E. g., BGA 28.5, et passim.

¹³ Cf. § 12.

¹⁴ Cf. §§ 14, 16 fin.

¹⁵ E.g., BGA 50.7, et al.

¹⁶ Although no conclusion for the Gs. can be drawn therefrom, the following statement made by Lanman, JAOS 10.339, is of interest here : ‘ There is hardly a single indubitable example of a voc. sg. neu. in the RV.‘ The Skt. neu. voc. in -a would, then, seem to be purely formal and paradigmatic.

¹⁷ Cf. gate, WSG § 339.

¹⁸ Being used as part of a masculine adj., the second member of this bahuvrihi compound forms its voc. like any other ă-stem.

¹⁹ Cf. notes for MS readings.

²⁰ Jn. Av. Gr. § 224.

²¹ My figures on the voc. do not include adjs. modifying nouns in the voc., nor nouns and pronouns in apposition with such nouns.

²² Cf. § 19.

²³ Cf. § 41.