iv. Origin of dependent pronouns: (1) Prefixed.—fy < Ar. *mene § 113 ii; dỿ ‘thy’ < Brit. *to(u) proclitic form of *tou̯e < Ar. *teu̯e;—ỿ ‘his’ < Ar. *esi̯ó: Skr. asyá; y ‘her’ < Ar. *esi̯ās: Skr. asyā́ḥ, § 75 vii (2);—an ‘our’, Bret. hon, hor, all for *anr, which (like Ir. ar n- for *anr n‑) represents regularly (§ 95 ii (3)) Kelt. *n̥sron < *n̥s-rōm: Goth. unsara, with suff. ‑(e)ro‑: cf. Lat. nostrum with suff. ‑t(e)ro‑;—ny before hun < *nes or *nos: Skr. naḥ acc., gen., dat.;—awch ‘your’, formed from chwi on the analogy of an: ni;—eu ‘their’, O.W. ou, Bret. ho, is probably for *w͡y unaccented, and so from *eison < Ar. *eisṓm: Skr. eṣā́m ‘their’ < *eisṓm, Osc. eisun‑k; for the weakening of unaccented w͡y to eu see § 78 iii;—yn, ych before numerals < *esnes, *esu̯es: Goth. izwis ‘you’ acc. < *esues;—yll is a form of an l-demonstrative § 165 vi, perhaps < acc. pl. *ollōs < *óli̯o- or *olno‑: Lat. ollus.
(2) Infixed.—Gen.— ‑m, ‑th see § ii (1); Brit. *men caused the rad. of tenues, the nas. of mediae § 107 iv, and as the latter was generalized for fy, the former was for ’m;— ‑e or ‑y is merely the prefixed y contracted with the preceding vowel;— ‑n, ‑ch are the prefixed forms with the vowel elided;— ‑e or ‑y ‘their’, originally only after o ‘from’ and *do ‘to’; thus oe or oy ‘from their’ < o *w͡y contracted; similarly the rarer oe ‘to their’; ay ‘and their, with their’ is formed on the analogy of oy, instead of the orig. ac eu which also survived, as oc eu ‘from their’ was formed on the analogy of the latter, instead of orig. oy (o ‘from’ had no ‑c);—i’w ‘to his’, etc., Ml. W. yw met. for *w͡y < *do ĭ ‘to his’ contracted after *esi̥ó ‘his’ had become *i̯, but early enough for *oi to become *w͡y, see § ii (1); the metathesis is actually attested in nuy (≡ nw͡y) > nyw, see below.
Acc. (dat.).— ‑m, ‑th < *mm‑, *tt‑ from acc. *me, *te, dat. *moi, *toi, originally used after the neg. ny, the tense part. ry, etc., which caused gemination of the initial; in Ir. also the forms after nī, ro, no, do, etc., are ‑mm‑, ‑t- (≡ tt); see § 217 iv (1); after the rel. a which causes lenition, ‑m, ‑th must be analogical; the rad. initial after ‑m is due to the analogy of ‑m gen.;— ‑n (Ir. ‑nn‑) < *nes, see § (1); ‑ch by analogy; the syllabic forms prob. developed thus: *pann m cl- > *pann m̥ cl- > pan ym clywai; so n > n̥ > ỿn; ỿth, ỿch by anal.; cf. heb ỿr § 198 iii; on the whole this is more probable than that ỿ- represents the vocalic ending of pann lost elsewhere, which is the explanation of the corresponding Ir. forms generally assumed (Thurneysen Gr. 246, Pedersen Gr. ii 145); in any case the y- is not the rel. y, which is not used after pan § 222 xi (2), so that the form pan y’m is misleading and wrong;— ‑e, ‑y, in ae, ay ‘who…him’, for ai *ĭ contracted; syllabic y < *ĭ; *ĭ < *en < *em ‘him’; the nasal ending caused the rad. of tenues, which was generalized; ‑s from the fem. acc. *sīm ‘her’, *si̯ās ‘them’, with the initial doubled as in *mm‑, *tt‑, so that it gives ‑s (not *h‑); in Ir. ‑s- is f. sg. only; in Corn. it is f. sg. and pl.; in W. extended to the m. because the m. *ĭ was lost after ni; thus *ni caf ef became nis caf ef on the anal. of nis caf hi; so ae ‘who…her’
