krïn̥ʹαχtə, ‘wheat’, O.Ir. cruithnecht; kʹɔ꞉ ·brin̥ʹə, ‘haze portending heat’, cp. Di. ceobhrán; Lʹin̥ʹə, compar. of Lʹαhən, ‘broad’; ræn̥ʹαχ, rαin̥ʹαχ, ‘fern’, M.Ir. raithnech. (c) < chn, e.g. dʹïn̥ʹu꞉r, ‘set of ten’, Di. deichneabhar, O.Ir. dechenbor; in̥ʹi꞉Nʹ, ‘brain’, M.Ir. inchind.
Note on the r sounds.
§ 265. Corresponding to the four l and four n sounds described above we expect to find four parallel r sounds (R, r, Rʹ, r), but it would seem that almost everywhere r has been generalised for R and there are no remains whatsoever of Rʹ. If any traces of Rʹ had survived we might expect to find them in such sandhi combinations as Nʹi꞉rʹ rʹïgər se, ‘he did not answer’; fʹαrəg ə Nʹirʹ rʹiʃNʹæʃi꞉ (fʹrʹiʃNʹæʃαχ), ‘the temper of the irritable man’. I have tried all imaginable combinations but have never heard anything but rʹ. It is easy to form a palatal r with the articulation of Lʹ, Nʹ, i.e. with the front rim of the tongue pressing against the lower teeth, but I cannot conceive why the sound was given up. Of R on the other hand there are distinct traces, but not in the initial position where it has become r. Fortunately Donegal has not lengthened the vowel in every case before R as Munster and Connaught have done and it is at the end of monosyllables after a short vowel that we must look for the sound.
9. R.
§ 266. By this symbol we denote a long, strongly trilled r with the tongue vibrating against the alveoles just above the upper teeth. Whether in pronouncing R the point of the tongue originally acted against the teeth as in the case of L, N, it is now impossible to say. O’Brien in his Grammar says (p. 11)꞉ “It (viz. r) is sometimes written double, as barrach tow; earr, champion; and is then strongly pronounced, with a longer dwelling on the sound of r than if it were written singly”. Now O’Brien was according to O’Donovan a native of Meath and it is quite likely that traces of R were preserved in that district as they have been in the west of Ulster. Quite unnecessarily Pedersen remarks (p. 38) that it is inconceivable that rr in barrach should be long. For traces of R in Scotland see Henderson, ZCP. iv 523.
