Bartholomae explains Av. ainita- “nicht vergewaltigt, - gekränkt” (Yt. 13.34; 51; 63) and ainiti- “milde Behandlung, Milde” from an-inita-, resp. *an-initi- (owing to haplology, see Altiran. Wörterb. 125 f.). This theory is the inevitable consequence of his positing an Avestan word [initi-] “Vergewaltigung, Kränkung”, supposed to be attested in the Instr. Sg. inti- Vend. 18.61 and in nitī. Y. 30.11b. If however [initio] had really been in use both in Gāthic and in Later Avestan, it would be hard to conceive how its opposite *an-initi- could have become [aniti-], since the Avestan speaker, in that case, must always have been aware of its origin as a mere negation of [initi-]. Besides, a participle *inita- would be anomalous in several respects. In the first place the present inaoiti [inauti], corresponding to Ved. inóti, must be referred to a root i-, as there is no indication to show that from the stem i-nau- a secondary root in- had been created in Indo-Iranian. The participle -inita-, only twice attested in a comparatively late Vedic text (úpenita- Šat. Br. III) is certainly an occasional new formation of Sanskrit. The Indo-Ir. noun Ved. énas-, Av. aēnah- “act of violence, crime, sin” contains the IE. suffix *-nes- and is derived direct from the root (cf. Ved. rékllas-,Av. raēxinah- from the root *raik-). As for Ved. iná- “strong, mighty, fierce”, there is no reason why it should rather be analysed as in-á- (e.g. Lindner, Altindische Nominalbildung, p. 33) than as i-ná-:. the nominal stem derived from the present stem is, indeed, -invá- (Wackernagel, Altind. Gramm., II, 1, p. 181). Secondly Bartholomae disregarded the fact that, while in Sanskrit participles in -ta- with an analogical “connecting -i-” are quite common, they are completely unknown in Old Iranian. This is apparently due to the circumstance that the IE. formation in *-H-tó- contained a consonantal laryngeal, which vanished in Iranian while being secondarily vocalized in the Indian branch. Thus to Ved. gbhītá- (with a specific Indian lengthening of i<H after a labial sonant) corresponds grpta- in Avestan. The mutilated Old Persian form agrbi ( … ) in Dar. Beh. 2.73, for which Bartholomae, Tolman, and Kent read āgarbīta “seized” (p. p., with a long i!), is more likely to stand for āgrbi<ya> (: Ved. aghyata). Accordingly Fravrtiš agrbi<ya> anayatā abiy mām means “F. was seized (and) led to me” (cf. 2.88). As far as the noun [initio] is concerned, the fact should be stressed that even in Sanskrit the “connecting -i-” is confined to the participles but does not, as a rule, occur in the verbal nouns in -ti-.