vocalic form. It was undoubtedly spoken r after all vowels then as now, except when a pause came between the words; for we find early examples of r even after diphthongs; thus kir llaw r eirccheid b.b. 10 ‘beside the suppliants’, mi yw r iarll w.m. 137 ‘I am the earl’, gwirẏon yw r vorwyn do. 138 ‘the maid is innocent’, erglyw r pobloeẟ r.p. 1201 ‘the peoples will hearken’. In some cases y is written where the metre requires r as Pa gur yw y porthawr? b.b. 94 ‘What man is the porter?’, where we should have yw r, as the line is 5 syll. Sometimes yr is written before a consonant: Pieu ir bet b.b. 66 for pieu’r beẟ ‘whose is the grave?’; llyma yr weẟ r.m. 2 for llyma’r weẟ ‘this is the manner’. In the early Mn. bards ’r is regular, esp. after pure vowels; and it is general in later prose, e.g. the 1620 Bible, though not without exception here. Pughe attempted to substitute y for it everywhere, and under his influence y was adopted in many late edns. of the Bible, except after a, o, i, na. This preference for y is chiefly due to the mistaken notion that r forms no part of the word, but was put in before vowels “for the sake of euphony”. We have seen above that the article is yr, and of the clipped forms ’r is older than y.
iv. The Ir. article is ind, after prepositions sind, from Kelt. *sendos, which gives W. hynn ‘this’, see § 164 vi. This occurs in W. in yn awr ‘now’, lit. ‘this hour’ (O. Bret. annaor, Ir. ind or sa), and y naill for *yn aill § 165 (Bret. ann eil § 166 iii, Ir. ind-ala). The art. in Corn. is en or an; in Ml. Bret. an; in Mn. Bret. ann before vowels, t‑, d‑, n- and h‑, al before l‑, ar before other consonants (so the Bret. indef. art. eunn, eul, eur, from un ‘one’).
Pedersen Gr. i 153 ff. quotes late examples of n > r after a cons. in Ir. dialects and Bret., and one or two cases of the change before a cons. as Ml. Ir. marbad for O.Ir. mainbad, Bret. mor-go ‘horse collar’ for *mon-go (obviously cases of dissim. of nasals). No such change as n > r is known in Welsh, which prefers to change r to the easier n § 100 i (2). W. yr can only be identified with Ir. ind by a rule made ad hoc; this is the only form of the art. in W. (yn awr is not ‘the hour’ but ‘this hour’); the ‑r abounds in the earliest period, and cannot be compared with Bret. ‑r, which is late, and may have spread from ar before r‑. The fact that there is a demonst. pron. ar in W. used before the rel., see § 164 v, makes the derivation of yr from hynn still less probable. There is no reason why the W. and Ir. articles should be the same word; the use of a demonst. as art. is much later than the separation of the P and Q groups. Gaulish has no art.; Pedersen Gr. ii 177 quotes σοσιν νεμητον ‘this temple’ as an example of the art. in Gaul., which is as if one were to quote in hoc tumulo from a Lat. inscr. as an example of the Latin “article” hic.
Though common in the O. W. glosses and prose fragments, the art. seldom occurs in the early poetry; it is not found in juv. sk., and is rare in the b.a.: Gwyr a aeth Gatraeth ‘[the] men who went to Catraeth’. It does not occur in O. Corn. or O. Bret., see Loth Voc.
