kimkat0354k Observations On The Welsh Pronouns. Max Nettlau, Ph.D. (Fiena, Ymerodraeth Awstria 1865 - Amsterdam, Yr Iseldiroedd 1944)


● kimkat0001 Yr Hafan

● ● kimkat2001k Y Fynedfa Gymraeg
● ● ● kimkat2194ke Cyfeirddalen yr Adran Ramadeg

● ● ● ● kimkat2194k Y Gyfeirddalen i Ysgrifa Max Nettlau

● ● ● ● ● kimkat0352k Y tudalen hwn



0003g_delw_baneri_cymru_catalonia_050111(delwedd 0003)






Gwefan Cymru-Catalonia
La Web de Gal
les i Catalunya
The Wales-Catalonia Website



VOL. XVIII. 1887. Pp.

Y Llyfr Ymwelwyr / El Llibre de Visitants / The Guestbook:

a-7000_kimkat1356kBeth syn newydd yn y wefan hon?


(delwedd 4665)

llythrennau cochion = testun heb ei gywiro



(delwedd B2694) (tudalen 113)

Y Cymmrodor.

Vol. VIII. Cared doeth yr Encilion. Part 2.




The following observations form a part of my dissertation Beitrge zur cymrischen Grammatik, the frst two chapters of which, containing the Introduction and the section on Vocalism, were printed a month ago in Leipsic. (1) I must refer my readers to this Introduction, quoted herein as Beitr, for more detailed accounts of the manuscripts and books cited below. I have collected from the sources available to me what materials I have been able to obtain relaitng to the history of the Welsh language and to its dialects, and a portion of these materials that, namely, relating to the Pronouns is comprised in this article. My chief aim has been to define the phonetic value of the orthographies found in manuscripts, and to separate archaisms from the analogical neologisms in which the modern language abounds. I should be very glad to see additions to the facts I give made by native Welshmen; they are principally needed with regard to the more accurate localisation of dialectal forms and to their phonetic description. I hope to have the opportunity of publishing further articles of similar character on other parts of Welsh Grammar.

(1) Beitrge zur cymrischen Grammatik, I, Einleitung und Vocalismus. Leipzig, Mrz April 1887, 79 pp., 8vo.



(delwedd B2695) (tudalen 114)


My observations are arranged in the order of the cor- responding parts of the Grammatica Celtica (pp. 368-409, ed. Ebel), to which they may be found to contain some additions.

A. Personal Pronouns.

[1.] My, tydi, etc, are formed by doubling the single pro- nouns; the accent may be on the first or on the second syllable, its position depending apparently on the kind of emphasis given to the pronoun. Griffith Eoberts {Gramm., 1567, p. 33[125]) and John Davies (Gramm., 1621) give examples of both accentuations. Spurrell (Gramm., ^81) has myf, etc. Hyhi, for hihi, is quoted by W. Morris in Addit. MS. 14,947, l 46, from the Anglesey dialect.

[2.] Y fi, y ti, y fe, y fo, y hi, y ni, y chwi, y nhw (y nhwy) are forms of the colloquial language. Cf. eb fi, eb y fi, in North Wales eb yr fi (Davies, Gramm., 1621, p. 136), eb yr di, eb yr ef, etc.; Addit, MS. 15,059, f. 209(x: eber hwnnw; f. 21()a, ehr Q,oeg{raith Gwgan). The article yr became so intimately connected with eb (heb), that the pres. sec. is not ebai yr fo, but ebrai fo (eb-(y)r-ai fo). This morphological deformity is not more to be wondered at than, e.g., the Lithuanian eikszte, .e., eik (the imp. of eimi, I go) sz(remains of szen, hither) te, the termination of the 2nd pers. pl. imp. (see Kurschat, Litanische Gram^natih, 1876, 1178). The singular eik-sz (eik szen), as in Welsh ebr (eb yr), was believed to be one word, and made accordingly the stem of verbal flection. Cf. Myvyrian rchaiology , 2nd edit., Preface, p. xxvi&: ebra f o, ebra hi, ebra nhw, with the a from final ai and e of the Yenedotian dialects. I may add from Addit. MS. 14,996, f 81 (1750): ebre Hugh Sion.

[3.] The forms of the 3rd sing. masc. of the pers. pronoun are one of the chief discernants of the northern and southern dialects. The efe, fe before verbs, of the literary language


(delwedd B2696) (tudalen 115)


and the southern dialects, is replaced in North Wales by mi.

Cf. Peter, in V Ci/mmrodoi\ i, p. 15, aiid the following examples, taken from letters in the Carnarvonshire dialect in Yr Arweinydd (Pwllheli, 1856-9): mi bydda i ar plant yma yn rhicy ddisgwil dy welad di yn dwad (dyfod) gartra; mi rouddwn i yn dy welad ti; ac mi rouddat ti yn ista ar ben rhicy (rhyw, see mj Bergc, 110) hen focs (box) yni hi; y rwsnos (wythnos) dwytha mi ddaru mi (i mi) fryuddwydio bryuddwyd digri iawn am dana ti; ac mi stopiodd y goits (coach) wth lidiart y ffor; mi rydan ni wedi disgwil llawer iawn; mi welwch lidiart hyuarn a trowch trw hono ac mi gewch etc. (26, 2, 57); ac mi ddoth AVil ar papur newudd yma ddou, ac mi ath Cadi i ddarllan o, ac mi rodd arni hi gwilidd, nes oedd hi bron a mynd ir dduar ar dy gywnt (account) di (30, 10, 56), etc; Hanes Caban f ewythr Tomos, 1853 (Merionethshire): mi alle hyny fod, p. 7, etc.

Mi can be nothing but the pers. pronoun of the Ist sing., ■which has by the operation of analogy exceeded its proper functions. It is not clear to me, as I do not know how old this use of mi is, whether there is any connection between the extension of mi over all the persons, and the fact of the South-Welsh pers. pronoun, 3rd sing. masc, efe, ef, fe, e, being totally discarded in North Wales, where efo, y fo, fo, o, are the only forms used since perhaps the sixteenth century. Even books written in the literary language may often be recognised as edited in North Wales or South Wales by using efo, fo, 0, or efe, fe, e, in preference. Efe is ef + ef, the final f being lost in pronunciation long since, and ouly written for reasons of etymology. Ef-o, I think, contains at the end the same pronominal element of demonstrative meaning as yco, yno, on which see Revue Celtique, vi, p. 57. J.D. Ehỳs(ŵam7?i., 1592, p. 65) writes efo, yf, f. Efo (he) has the accent on the second syllable (D. S. Evans); so the unstressed e was lost (fo), as sef sprang from ys-ef. was abstracted from efo, since e (ef) existed besides efe (ef-ef). In the Preface to Llyfr Goeddi Gyffrcdin, 1586 (on which see my Bcitr., 12, 3) efo, fo are said to be North-Welsh. Cf., e.g., Addit. MS.



(delwedd B2697) (tudalen 116)


31,056, f. 17:yr ysgol fom gyrwyd {Eanes y Trwstan, Powys), etc. Enclitic fe becomes fa in the eastern Gwentian dialects, like all es in final syllables; cf. Y Tyioysog or Crymraes (Llanelli): onite, ond do fa do; Y Bedyddiwr (Cardiff): on te f a; Y Geninen, vol. iii (see Beitr., p. 26): ontefa, in East Glamorganshire.

[4.] There exists another efo, with the accent on the first syllable, used only in the North-Welsh dialects, which has the meaning of with, by means of. I am not acquainted with any Yenedotian manuscripts written in popular language prior to the eighteenth century, so the oldest occurrence of this word to which I can refer is the following, in a manuscript of Lewis Morris. He writes (in Addit. MS. 14,923, f. 134): Efo g ef,efo g efo,is common in Anglesey, for which they are laughed at by other countries; Engl. with him.

Cf. Hope, Cyfall iW Cymro, 1765 (Powys), p. 4: a chwery hefo chwe bys (by Dafydd Jones o Lanfair Talhaearn); Hughes, An Essay oii the .... State of the Welsh Language, 1822 (see Beitr.^ p. 26): N.W. efo ni = S.W. gyd ni; F Traethodydd, iii, p. 9 (in Amrywieithoedd y Gymraeg by Morris DaTes): N.W. gydago, efo ag ethi hi, gydag efo, efo fo, efo ag efo, efo ag o, efo ag ethyn nhw, efo nhwthe; in Yr Arweinydd: achos i fod o yn rindia gyrwinol hefo yr hogan W. acw, 17, 7, 56, hefogo 11, 12, 56; hefo mi, hefoti, etc; Cabanfewythr Tonios: hefo phethe wel hyn, p. 23, hefo th di, p. 473, etc.

fo is very often written hefo, h being due to the stress on the first syllable, as, e.g., hyny is nearly always written for yny in the parts recently edited from Hengwrt MS. 59 (Eevue Celt., vii, 4). If fo were the pronoun of the 3rd sing. masc, the assumption of a paratactic construction would be required to explain a sentence like ef aeth efo mi (he weut he I; he and I; he with me), and froni such sentences the use of fo must have been extended by analogy to the other persons. As I am not aware how old this prepositional use of fo is, I ought,


(delwedd B2698) (tudalen 117)

OBSERYATIONS ON THE WELSH PRONOUNS. 117 perhaps, to abstaiu from furtlier conjectures, but I may say, that such an elementary parataxis seems to me to be not very probable, and I venture to suggest another explanation of fo altogether. [5.] I quoted above: hefo phethe wel hyn. The aspiration of l could be explained by fo having completely tal<en over the functionof ac, a(with) , and therefore causing the aspiration of tenues. But there exists also an older form of the pre- position ac, a (Ir. oc), namely, oc, o, which causes aspiration of initial tenues. Powel says, in Y Cymmr., vol. iii, o is used iu the dialect of Glamorganshire, to denote the instrument. Cf. Seren Gomer, vol. i (1814), No. 19, where the following Glamorganshire expressions are given: codi glo or rhaw (r rhaw), tori cnau ou dannedd, taro ci o asgwrn, ffustio haiarn or ordd, etc. In Skenes Taliessin, 17, is printed: ny thy ogyfeirch o chwynogyon (that thou shalt not be addressed by vulgar oues?). Cott., Tit. D. 22, f . 4: o chledyfeu (see Powel, Y Cymmr., iii). The most numerous examples I found in the Gwentian manuscript, Addit. 14,921, containing a frag- ment of a Welsh translation of Sir Jolin Maundevilles Travels, 16th century: o cheric (with stones), f. 18; toi o lwm, f. 19, f. 20 (toi o blwm, f. 15&); o than yffernol, f. 28; thom keffyle, f. 40&; dal o gwaetgwn, f. 59; ac y laddson y mydwy or kleddy, f. 46; Uad . . . . or kleddy, f. 21; gwielen . . . . or hwn y rannoedd ef (Moses) y mor koch, f. 21; ac or weilen {sic) hono y gneth ef llawer o ryfeddody, f. 21; (in the same manuscript oc aud o, from, are used: gwedy y gwitho oc ayr ac arin, f. 62; wedy gnythyr yniU o ayr ar nU oc arin, f. 62, etc). Oc, o, from, has become c, , in the modcrn language, where it is con- nected with the foUowing relative, : Middle Welsh oc a, ar a; and r (written or a, ar ), is now ag , r . On other instances of o becoming a in Welsh, see Beitr., 55. [6.] I think in fo (with) to be this older form of ac, a.


(delwedd B2699) (tudalen 118)


of wliich I just now gave examples. If ef- in fo is the pro- noun he, the further extension of an he with could much more easily be accounted for than the extension of a simple paratactic he. But I rather think this ef- to be a pro- nominal element of a somewhat faded demonstrative mean- ing, which also occurs in hefyd (Middle Welsh, heuyt). On -yt see Bevue Celtique, vi, p. 57, seq. There are, however, Breton forms of hefyd, presenting further diffculties. At any rate, the common usage of ag with efo (efo ag) cannot form an argument of any value against these suggestions, unless its earlier occurrence could be proved, as ag may be here of very modern origin, caused by efo (and o, with) becoming obscure, since oc, o (with) was supplanted by ac, a.

[7.] Wynteu contains wy + nt, the termination of the 3rd pers. pl. of the verb (as Ir. iat), -i- eu, the plural termina- tion of the u-stems; ynteu took by analogy -nt- and -eu from wynteu. There occurs also yntef, containing the pers. pro- noun, 3rd sing. masc, ef; and this -ef was spread over all the other persons: minnef, tithef, etc.

Cf. Owen, Ancient Laws, 1841 (fol.), p. 528, MS. a: o gwnaent wyntef; Salesbury, New Test. 1567: yntef, mynef, minef, f. 274; mal tuhun (tithaw {sic\ tithau, tithef), on the marg., f. 36; Y Drych Christianogaicl, 1585: y gallwn innef, Cl, ninef, f. 19a; J. D. Rhys, Gramm., 1592: minnef, tithef, etc.; Hom., 1606 (Pre- gethaii .... gwedi eu troi ir iath Gymeraeg, drwy waith Edward James): yntef i, p. 104, etc.

I cannot ascertain whether such forms were really spoken, or whether they show only an etymological orthography, since both ynteu and yntef were pronounced as early as in the 16th century ynte (or ynta). It is not altogether im- probable that yntef had a real existence, as in the modern Yenedotian dialect another combination of ynteu and ef is in common use, namely, ef ynteu, pronounced fynta. Cf Yr Arweinydd: wrthoch chi a fynta, 17, 7, 56; medda fynta.


(delwedd B2700) (tudalen 119)


cysdlad a fynta, 28, 5, 57; iddo fynta, 11, 12, 56; Sweet, Spoken North Welsh, p. 441: nta, fnta (from Gwynant).

[8.] Hithi (she) seems to be a rare form, which I only found mentioned by L. Morris in Addit. MS. 14,923, f. 134b: hithi and nhwthw, she and they, in Merionethshire. It is, of course, a combination of hitheu and hihi, hi, as nhwthw is of ynthwytheu and ynthwy, hwy.

[9.] The pers. pronoun chwi, chwichwi, chwychwi, chwitheu (you), seems to be excepted from the well-known South-Welsh change of chw into hw, wh, as the colloquial forms used all over Wales are chi, chichi, chitheu. The oldest examples I know are printed from Llyfr Gwyn Rhydderch (E. Williams, Hengwrt MSS., vol. ii): a mae awch kyghor awchchi wrth hynny, p. 158; o honawchi, p. 153 (hwchwitheu? p. 235). From the sixteenth century downwards: Salesbury, New Test.: os profasochi, byddwchithay, f. 352b; ynochi, f. 344a; cenychi, f. 348a; Add. MS. 15,986 (sixteenth century, see my Beitr., p. 19): chwchi, f. 14a, 15a; chi, f. 13b; Y Drych Christ., 1585: chwchwi, etc. In modern dialects: Dimet.: a ody chi, Seren Cymru, i (1856-7), p. 429; Gwent.: i chi (ydych chwi), oe chi, chi, Y Tyw. ar Gymraes; Venedot.: beth ydach chi, Yr Arw.; Powys.: mi rydach chi, Cab. fewythr T., p, 19, etc. Hwchwitheu is of very doubtful existence; the scribe may have looked first only at the initial chw, for which he wrote his hw, but afterwards seeing that the word was chwitheu, which he pronounced chitheu, he wrote chw, as he nearly always does in this word. I think that chi for chwi took its origin in the position of chwi after the 2nd pers. plur. of the verb, and in prepositions with a suffixed pronoun of the 2nd pers. plur. At the time of the other chw giving way to hw, wh, the chw of chwi was continually influenced and supported by the ch of the verbal termination, and thus conserved. Afterwards w was dropped in the unstressed forms by a phonetic law, of which I do not know other


(delwedd B2701) (tudalen 120)


examples common to all dialects; but chefrol (chwefror), chedy (gwedy), cheugain, damchein, etc, are occasionally found.

[10.] The forms of the pers. pronoun, 3rd pers. plur., are very numerous, a phenomenon brought about by the analogical influence of the pronouns of other persons, by the easy change of the position of the several, still very transparent, parts of it, and by their pleonastic doubling.

On wy-nt-eu, see 7. From this form -teu was abstracted and transferred to the pronouns of the other persons, where, by the influence of some fnal consonant, th in -theu was caused. It is almost certain that -theu originated in the 2nd pers. sing. *T-tu (the quantity of n in this pronoun depends on the stress laid upon the form) -f- tou (*-toves) seems to have become, by dropping the unstressed , *tt-tou, hence titheu. So *mmi-tou gave *min-tou, hence minneu. The pers. pronouns, Ist and 2nd pers. pL, are believed to have ended in -s (as nos, vos); so it is necessary to assume that ninneu and chwitheu are formed after the analogy of the respective sing. forms minneu, titheu. Finally, -theu was introduced into the 3rd pers. sing. fem. (hitheu), and the 3rd pers. plur. hwytheu.

[11.] Hwy-nt + hwy gave hwyntwy (hwntw); hwyntwy was made yntwy by the influence of the sing. ynteu, or perhaps yntef, as ynt-wy and ynt-ef are exactly parallel forms (yntwy, ynhwy, ynhw, nhw). Still more complicated are nhwynte (y-nt-wy + nt-}-eu), nhwthau (y-nt-wy-|-thau), nhwthw(y-nt-hwy + th[au] -I- wy) and wyntwthe (wy-nt-wy + thau).

Cf. Owen, Lmos, MS. A.: vintoe, p. 74; Cott., Tit. D. 22, f. 6b: 6nteu; Salesbury, N. Test.: wyntwy, wytbeu, gantbyn bwytheu; (R. Davies): y mayntw f. 329&; wrtbyntw f. 342a; bwyntw f. 3300, wyntwtbe f. 335, bwynt, wytbe a wuant {at y C.) (Huet): oe geneye yntbwy f. 385; Gr. Roberts, Gramm. p. 76 gwelantwy, gwelanbwy, gwelant yntwy ne ynnbwy ne bwynt ynbwytbau, yntbwytbau, bwyutau, p. 31 (233), nbwytbau, p. (231)


(delwedd B2702) (tudalen 121)


Athi: Grst., 1568: cans, yiihwy a feddianan yddaear, p. 55, cans hwyutwy pie teyrnas uefoed, p. 55, cans ynhwy a gan, p. 56, nhwy (often); Addit. MS. 14,986, 16th cent.: nhwy, f. 376; North-W. Addit. MS. 31,056, f. 20a: nanbwythe; Addit. MS. 14,898, f. 78: a nhwythai, f. 79a; Addit. MS. 15,(i59, f. 209a: ond nhw ni welson moi gilydd erioed; Cyfaill ir Cymro, 1765: nhwythau, nhwythe,p. 8. In the modern language: ynhwy, ynhẃ, nhw (Rhŷs, Lectures, ^p. 55); y nhw, y nhwy, nhwthau, nhwythau (Rowlands, Gramm., *p. 49); nhwthw in Merionethshire (L. ]\lorris, s. 8); South-W. nhwy North-W. nhw (Rowlands, Wehh Eiercines, 1870); Y Traethodydd, iii, p. 11, has: S. W. hwy, hwynt, hwyntwy N.W. nhw, nhwy, nhwythau.i In the SouthWelsh (Gwentian) MS. of Hanes Grujfydd ab Cynan, in J\Jyv. Ach., occurs several times wynteu, whilst hwytheu is given in the notes from the North-Welsh MS. (^p. 724, 730, 731).

Cf. Dimetian: u bod nw, Ser. C, i, p. 212, a nhwynte, iii, p. 324, etc.; Gwentian: gyda nhwy, gentu nhw, Y Tyic. aV Gymr., I, p. 134; a nhwynte, ii, p. 95; Venedotiau: nhwthau, nwtha, Yr Arw.; Powys.: u bod nhw, nhwthe, Cab.f ew. Tom.., pp. 53, 60: o honyn nhwthe, p. 62.

1 In y Traeth, l.c, the fact is mentioned, that the South-Welshmen are nicknamed by their northern neighbours, on account of this use of hwyntwy, Hwyntwys, and hwyntwy bach yn awr (S. W. nawr = N. W. rŵan). In a letter, sigued Rhobin Ddu, in Addit. MS. 15,030, f. I30a, gan Hwyntwy is used to denote South-Welshmen. lolo Morganwg, in Addit. 15,027, f. 79a, calls the North-Welshmen Deudneudwyr (from the infin. deyd, gueyd), and the South-Welshmen Hwyntwyr. The following, containing.other such names, is an extract from one of the articles written in the colloquial language in Seren Cymru (Carmarthen). A uative of Merionethshire is iutroduced, speaking his own dialect: ma nw yn y galw I yn hen Northman fel tase nw yn galw hen leider arna I; ac ma nw yn galw hen Hiontws a Gwentws a phethe felly yu y Gogledd ar wŷr y De; . . . ma llawer o drigolion Mynwy a Morganwg yn gas filen i bobol Sir Gaerfyrddin, pan ma rheiny yn dod ir gweithfeydd; ma nw yu u galw nw yn Hen Shirgars a Heti Gardies, Shacld neici ffam a rhyw hen lolos felly. Cf. also Punch Cymraeg, No. 29 (Ebbw Vale): haid o Gardies. Newi ib the Dimetian form of newydd.


(delwedd B2703) (tudalen 122)


B. SuFFixED Personal Pronouns.

[12.] The manner in whicli terminations similar to those of the verb are suffixed to the preposition in Welsh to express a pers. pronoun clependent on it is to a great extent common to the Welsh and the two other Brythonic languages; cf. Corn. ynnof. ragof genef, worthyf; W. ynof, rhagof gennyf, wrthyf, etc. This shows the distribution of the three groups -of, -af, -yf to be pre-Cymric, and therefore it cannot be discussed here, where I propose to limit my remarks to the history of the Welsh language only. The history of these suffixed terminations in Welsh is a very complicated one, there being numerous shiftings from one series into the other, though certain rules governing these changes may be ascertained.

I shall exclude provisionally oc, do, aud also eiddof, which require some separate remarks. All the forms quoted in the following general remarks will be found exemplified subsequently under the head of the respective prepositions.

[13.] Davies (Gramm.) gives three series of pronominal terminations: -of -ot -om -och -ynt (er, yn, rhag, rliwng, oc[ohon-], tros, trwy, heb, hyd, is, uch; eiddof), -af -at -om -och -ynt (ar, at, tan), and -yf -yt -ym -ych -ynt (can, wrth). AU prepositions may follow, in the Ist pers. sing. and 3rd pers. plur. principally, a fourth series: -wyf -wynt. The terminations -wyf of the old optative and -of of the old conjunctive were both spread over all persons, the extension of -wyf being promoted by its external identity with the verb subst. wyf. This latter fact caused -yw to be suffixed to the ord pers. sing. masc. (amdanwyf, amdenyw, like henyw, cenyw, from hanfod, canfod). The 2nd pers. plur. arnywch, wrthywch, cenywch, may also contain the 2ud pers. plur. of wyf, originated like amdenyw; it is also possible that the coincidence of -ym -ych -ynt in wrthym, etc, with ym, ych


(delwedd B2704) (tudalen 123)


and ywch, ynt of the verb. subst., was the source of this -y wch with a preposition (wrthywch, etc). There are in general scarcely any forms in Welsh grammar so open to the influence of numerous analogies as these prepositions with suffixed personal pronouns. It must, however, be pointed out that no transgressions into tbe series -yf occur, whilst the transgression of the -yf-series into the two others, and the mutual changes of -of and -af, are very common.

[14.] I found a curious example of the vast influence of verbal terminations upon these suffixed pronouns in Trijsorfa Gicyljodaeth neu Eurgrawn Cymraeg, Carmarthen, 1770. In this periodical a letter to the editor, signed Dafydd ap Ehisiart, is printed (reprinted in Y Tractliodydd, iv, p. 458), in which the foUowing dialectal forms from various parts of Wales are cited: ynnost, attast, iw eu gilydd, iw eu dwylaw, etc, allanol (oddiallan), tufewnol (oddifewn), hyd nod (hyd yn noed), abergofi (ebargofi), or braidd (braidd or o fraidd), yn eu holau (yn eu hol). I am not able to test the truth of niost of these statements, but iw eu gilydd (for example) may also, though very seldom, be found elsewhere. If these statements be right, ynnost and attast for ynot and attat are evidently formed after the analogy of buost, gwneuthost, etc; attast is to be held for a contamination of attatt and ynnost, though examples of doethast occur in various independent sources. I have nowhere found an example of these or similar forms, the admission of which, however, can do no harm, as they are neologies easily and fully explained.

[15.] Either these pronominal terminations seem to be directly suffixed to the prepositions, the vowels of some of whicli are altered on account of the earlier position of the accent (as rhyngof, and wrthyf from wyrthyf) or of certain infecting terminations (arnei: erni), or -n-, t, th, d, dd, -add- (-odd-?) are apparently inserted (arnof, rhagddaw, onaddynt). On -n- see below. As to the dentals, I am aware of the


(delwedd B2705) (tudalen 124)


explanation of theni given by Wh. Stokes (Kuhns Eeitschrift, xxviii, p. 97). s this is a very serious question, let me point out here the opinion which I helcl on these forms before seeing Dr. Stokess Olcl Irish Verb Sustantẁe. The dentals evidently spread froni the 3rd pers. sing. and plur., where -taw, -tu (see below) are instances of cases of the pronominal stem to-. The different kinds of dentals (t, th, d, dd) are the results of the meeting of -t with the diferent final sounds of the prepositions. The details of this matter are e^tremel}- dificult, and the Breton and Cornish should be compared in each case. In Welsh grammars most of the prepositions are said to use more than one intermediate dental (yndo, yntlio, ynddo, etc). Torms without dentals are aLso of frequent occurrence, especially in the modern language (gino, etc). These seem to follow forms like attaf, where, the final consonant being a dental, the i of -taw, -tu apparently does not exist.

[16.] The four prepositions oc, ar, tan, at, form, as is known, besides ohonunt, arnunt, y danunt, attunt, which follow the analogy of the other prepositions, onaddunt, arnaddunt, etc. There occur also, but very seldom, forms with -odd- (o noddynt, arnoddynt). These forms are likely to be considered clerical errors; but the Cornish anotho, warnozo, tend to support them, and I will therefore give here those which I have met with: Addit. MS. 22,356, a fifteenth century Cardiganshire manuscript of the Dimetian Code of Howels Laws, etc, styled S. in Owens edition (see my Beitr., p. 13): o nodynt, f. 77&, 80; y bob un o nodynt,f. 112, 120, 121 (o nadynt, Owen, p. 594; o honynt, ih.; o anadynt, p. 592 = f. 76); Salesbury, N. Test.: o hanoddynt, f. 152; arnoddynt, f. 212. S. has numerous other peculiarities of language, which are seldom met witli in manuscripts before the sixteenth century, as oy for eu in amhoyys, dadloeoed (cf. my Beitr., 96, where yn Thehoibarth, Lbjfr Acliau, p. 24, may


(delwedd B2706) (tudalen 125)


be added), rhoc (rhag, see l.c, p. 7fi), etc. Salesbury also iises many dialectal forms on one or two occasions, as weyneb (wyueb), etc. All these forms (-add-) are pre-Cymric (cf. Coru. auotho, warnozo, sing.; aueze, warneze, plur.; Bret. warnezo, etc), except attaddunt and y danadduut. I should suggest, that in a pre-Cymric period *on-tu, *aru-tu^ (to use later Welsh forms as types) were, from some reason not traceable by me, uot formed. Perhaps the 11 has something to do with this matter. Ou-o-tu, arn a-tu (or rather on-o-du, arn-a-du) have the of ohonof, the a of arnaf. This means, in other terms, that ohonof was conjugated (to risk this expression), ohonof, ohono-t; and thence ohono-daw. The Ist and 2nd pers. sing. aud plur. influenced the 3rd pers., as, conversely, iu most other instances the 3rd pers. influenced the others, by spreading their dentals (t, th, d, dd) over them (rhagddof, rhagof, etc). Both these effects of analogy are partly pre-Cymric.

[17.] A few forms require some further consideration. The 3rd pers. sing. fem. of many, if not of all, prepositions ends iu -ei and -i, which have been identified with the termiuatious of the pres. sec. -ei and -i. Ou the rare -i see E. Evans, Stud. in Cymr. Philology (Arch. Camhr., 1874), 26, and Ehŷs, Iiev. Celt., vi. and a are altered by the following -i into e. These doublets are: ohonei, etc, and oheni, recdi, etti, erni, genthi, amdeui. U has beeu trausferred by aualogy to other persons, as ohenynt, ernynt, gentho, etc.

[18.] The oldest form of the 3rd pers. plur. common in Cornish and Breton ends in -u (-tu). A few exaniples exist in early Middle Welsh manuscripts (ohonu, ganthu, racdu, etc). -nt of the 3rd plur. of the verb, as in wy-nt, and iu later Cornish warnothans, was suffixed to -u at an early period (-uut). In modern Welsh -ynt is usual. SincC; in the

Oc and ar have both the somewhat obscure?i, which I thiuk to be in both of the same origin; see below.


(delwedd B2707) (tudalen 126)


time of -unt, *-uni *-ucli never occur, I suppose -unt to have been altered into -ynt by the influence of the pres. sec- (-ym -ych -ynt, hence wrthym -ych; wrthynt).

[19.] There are forms of the 3rd pers. plur. in -udd (ganthudd, arnaddudd, etc.) given in the grammars, and said to occur in early poets; I think they are first found in Daviess Dictionary. I have no idea as to their explanation, nor do I know whether they exist at all. Information on this point could certainly be given from the unedited early poets in the Red Booh of Hergest. As t is sometimes written in eaiiy Middle Welsh for dd {Booh o Carmarthen) , and the line over a ii denoting n may be wrongly omitted, or have vanished or be oveiiooked, a transcriber may be supposed to have written ganthudd for ganthu(n)t, etc. There are scarcely any forms in Welsh grammar more puzzling to me than these. See, however, 37.

[20.] The following quotations are disposed under the heads of the various prepositions, arranged alphabetically in the three groups: -of, -af, -yf; concluding with oc, do, eiddof.

[21.] L, -of. Ur: J. D. Ehỳs {Gramm., 1592, p. 14) gives er duw, yr duw; er duw, ir duw, er, ir, and yr ieu, as dialectal forms. I am not able to localise these forms; I only found in Han. Qr. ah Cyn., yr hynny, ir, in the Southern, and er hynny, er, in the Northern manuscript, as printed in Myv., Arch., 2p. 7350, 725a.

Erof, ZZ. Gw. Rhydd., p. 109 (E. Williams); erom uine, Sal., N. Test., f. 227; Addit. MS. 14,986: erochwi, f. 96; erochi, f. 25; eroch, f. 265; Y Seint Greal: yrdow, p. 20; Addit. MS. 14,973 (1640, Eees Prichard): erddom, f. 75; also Cann. y Cymry, ed. 1672, p. 81: erddod, p. 116 (erod, p. 254); erddot ti dy han, Barddas, i, p. 318; erddwyf, erwyf, erof, L. Morris, Addit. MS. 14,944, f. 81; erof, etc, erddof, etc, Eowlands, Gramm., *p. 122.


(delwedd B2708) (tudalen 127)


[22.] Hcb: hebof or liebwyf (Eowlands, Gramm., *p. 121); hebdaw, etc. In Sal., N. Test., f. 341, heibom ninay is printed. If this form can be trusted, it is due to heibio, ]ike is, adv. iso, isof; uch, adv, ucho, uchof.

[23.] Hydot, etc, hydddo (Davies, Gramm.), hydddof, etc, isof, etc, require no further remark.

[24.] Najn: uemo, except of me (Spurrell, Dict.); nemof, -ot, -o, -i, etc, are given in D. S. Evans, Llythyriaetli. Cf. nam, named (but, since, except, Sp., Dict.), and Bret. nameit, meit, meiton, meitous, meitou, etc, ( Vann.; Guillaume, Gramm., 1836), nemedouf, etc (Zeuss, ^p. 728). I cannot explain e. There are amyn, yn amyn, and namyn. Either namyn can be explained as being (y)n-amyn cf. mae, sef (ym-ae, ys-ef); also probably nachaf,nychaf (yn achaf), or amyn sprang from namyn, as hoeth from noeth, and also in a few instances initial h from rh (by a wrong separation between, e.g., the preposition yn and the article yr, and the initial consonant of the following word). The forms of other Celtic languages must be compared here. Namwyn, of which many instances exist (MS. A of the Yenedotian Code: namuin, p. 58, namuyn, p. 59, etc [also namun, p. 58 = namwn]; Booh of Carm.: namuin, No. 5, 17; B. of. Hcrg.: namwyn, Skene, p. 249, col. 1031; namyn, col. 1186; Ll. Gw. Bhydd.: namwyn, p. 135), enhances the difficulties of this matter. I shall give the materials I have collected iu my article on the adverbs.

[25.] Bhac: rhoc for rhac occurs very seldom in Middle Welsh manuscripts. In rny Beitr., p. 78, I was able to give eight medieval and oue later example, viz., rocdi, L. (Dimet. Code,ed. Owen),p. 255; also r. = Harl. 958, f. 41i; ^Addit. MS., 22,356, f. 28.; roc S., f. 45, 100; rocdut, f. 45; Tit. D. 22: rog, f. 7dh; rogot ti, f. 167; also in the later MS. 15,038, f. 815: rogddo. All these MSS. are written in the Dimetian dialect, which is peculiarly couspicuous in S. and


(delwedd B2709) (tudalen 128)


in the Cottonian manuscript. To tliese forms, which I know, add from Bevuc Celiiqiie, vii, p. 427, rocdi (in the fragment of Ger. ab Erlin from Hengwrt MS. 59), and rogom ni, ih., p. 426, quoted from Yst. de arolo Magno, which I had not used (see Beitr., p. 15). I think rhoc is the older South- Welsh form of rhac. Cf. Gwentian, oc, o (Instrum., see above, 5), S. W. noc (nac), oc a: ag a. Ehoc contains rho and oc; on rho see below, rho rhwng.

Eecdi: B. ofTal., Skene,No. 14; Cott., O/eoj?. B.5, f. 2320 1 {Dares Bhrygius); f. 18Sa: raghdunt. Eecdi occurs also in Hengwrt MS. 59 {Bevue Ceique, vii). Salesbury, N. Test.: rhacddynt wy, f. 117&; rhacddwynt, f. 79&, etc.

[26.] Bho: Davies (Gramm.), Eichards, etc, give rhf, y rhof, poet. = rhyngof; rhf, rht, rhydaw, rhydi, rhm, rhch, rhydunt. Ehydaw, -i, -unt, occur very often in Middle South-Welsh manuscripts, where the North-Welsh ones have rhyngtaw, etc.

Cf . L. (Dimetian Code), Owen, pp. 235, 241, 252: 258: y rydunt; y ryngthunt in J.; V. = Harl. MS. 4353 (Gwent.), f. 37b: y ryda ynteu; W. = Clenp. A. 14, f. 35a: ỳ rỳda; f. 55: rỳdunt; Addit. MS. 19,709, f. 83: y rydunt; Hengwrt MS. No. 59: y rydi (Revue Celt., vii).

A few instances of the other persons are: er rot ath haul (Yenedot. MS. G., Owen, p. 458) = ỳ rỳghot tỳ ath dlỳet in MS. B.; y rhom ni ac wyntwy, Salesbury, iV. T., f. 196a, etc. Y ryoch, in Ll. Gw. Bhydd., p. 62, is formed by analogy, taking its ry- from the 3rd person.

InZeuss, Gr. Celt^ the forms of rho and those of er, yr, are mixed up. On p. 670, yrof i a duw is said to contain er, yr. This formula, given by J. D. Ehŷs {Gramm., p. 106) as rhofi a duw, and often thus written in mediaeval manuscripts, is correctly explained by Ehŷs, Y Cymmr., vii. Comparison with the Cornish re dev an tas, re synt iouyn, rom laute (Zeuss, -p. 666), makes this evident.


(delwedd B2710) (tudalen 129)


[27.] Rhwng. This preposition is a compound of rlio and wng. In Spurrells Dict. I find (also in the English-Welsh part, s. V. near) wng, preposition, near; adverb, here; which I have not met with elsewhere. In Dosparth Edeyrn, 981 (Williams), yngo (formed like iso, ucho), hard by, is given; cf. also n immwngc, explained in the margin disymwth, Gann. y Oymry, 1672, pp. 379, 400, 569.

In Addit. MS. 22,356 {=S., Owen, Laios), Cardiganshire, rong occurs several times: rong, f. 60a (=Owen, p. 195); y gwan (gwahan) yssyd rog mach d. a mach k., f. 73; r6g rog (=rong) r6g, f. 11 5a, 11. 19, 20. These forms, combined with others in the same manuscript o h6nynt, honn6, 116fryd, etc. (see my Beitr., p. 72, note 39) suggest some dialectal phonetic change, whicli brought about a close similarity in sound of o and w. It would be important for the localisation of this manuscript (see Beitr., p. 13), to know some details of this pronunciation from modern dialects. We for oe (see /. c, 80) is used in a part of Dyfed.

Some examples of rhwng with suffixed pronoun are: Eecl B. of Hcrg., y rygtunt, col. 558, 741; y ryngtunt, col. 730, 769 (rydunt, col. 567); ryngthi, col. 791; Ll. Gw. Rhycld.: y rygtunt, p. 3; y rygktunt, p. 315, 319; Calicj. K.V?> {=C., Owen): er rgth6nt, regth6nt, f. 177; Cleop. B. 5: iỳgthuht, f. 22; y rnthunt, f. 20; rwngthunt ar Saesson, p. 97; y rygthunt, f. 224 1; y ryghunt, f. 243 1. Jes. Coll., 141 {sQQ, Beitr.,-^. 14): rryngtho, f. 49; y rrynthvnt, f. 51; rryngthvnt, f. 45; rryngvnt, f. 54&. Sal., Neio TeM.: rhyngtwynt, f. 126&. Addit. MS. 31,050, rwngthynt, f. 6; Addit. MS. 15,059: rhynthyn nhw, f. 209, etc.

Davies, Gramm.: rhyngof, -ot, rhyngtho, rhyngddo; Pughe, Gramm., has rhyngwyf, rhyngthwyf, rhyngof, rhyngthof, -om, -ddom, -thom, -tom, etc.

In modern dialects: rhyngty nw, Scr. Gi/mni (Dimet.), i, p. 449; rhyngti nw, iii, p. 45; rhwutw i, Y Giceithiwr

VOL. vin. K


(delwedd B2711) (tudalen 130)


(berdare), 17, 9, 1S59; rhjngo i a B., Yr Ariv. (Pwllheli), 17, 7, 56; rhyngon, 28, 5, 57; Sweet, SiJohen North Welsli, p. 450, gives rhwngfi, rhngddoti, rhngddofo, rhangddoni, -chi, rhwng chi, rhngddynhw, from Gwynant.

[28.] Tros: In a modern Powysian text, Caban fcv-ythr Tomos, trost occurs several times: trost y dyn ene, p. 18; trost ben, p. 93; trost i ben nhw, etc. In P>reton exist dreist (Lon., Troude), drs, and drest (A^ann., Guillanme, Gramm., p. 90). I thinh these forms in t are wrong abstractions from trost-of, besides tros-of, t being the regular inserted dental.

Cf. Addit. MS. 14,931: dorstau, f. 52a (also in .S; = 22,356: ny thelir dim dorsta, f. 7b; drsta, f. 8a); trosda, f. Idh; drostau, f. 21?; (cf. twrstan trwstan, etc), Tit. D. 22: drossom ni, f. I8h; B. ofHerg.: drossof i.col. 755; drosda6,col. 751; Sal., N. T.: drosdi (amdenei), drosdei, f. 119h; trosoch eich hunain, f. 127&, drosdynt, f. 1166; drostwynt, f. 183a; drosswyf, droswi, Cann. y Cymry.

[29.] Triy: Cf. Clcop. B. 5: trwof vi, f. 38 (drwod, Ll. Giu. Rhyclcl, p. 139). Sal, N. T.: trwyddhei, trwyddom ni, f. 265&; trwyddwynt, f. 195, etc.; Athrav. Grist.: drwyddyn nhw, p. 36. trwddi hi, Cah.feiu. T, p. 38.

[30.] Uch: odduchof, odduchtof, BosjJ. Ecleyrn (WiUiams), 940, etc.

[31.] Yn: Sal,iV.r.: ynthaw, f. 135; ynthei,f. 3a; ynthynt, f. 00; ynthwynt, f. 12; ynoch eich unain, ynochwi, f. 1465; ynochi, f. 344; Addit. MS. 14,986, f. 8&: ynto. Davies, Gramm.: ynof, -ot, yntho, ynddo, poet. ynto, yndo; Williams, DosjJ. Ed., 940: ynddof, ynthof, yntof. In modern dialects: ynwyf, Ser. Cymru, i, p. 175, Y Tyio. ar G, i, p. 114 (literary language); ynthw i, Cah. fciu. T, p. 338; yntho fo, pp. 7, 252; ynthi hi, p. 252; yno fo, Yr Ariu., 17, 7, 56, yni hi ib; cf. gyno fo, in the same dialect.

[32.] II. -af. Ar: The third personplural isgiveninDavies Dictionary as aruaddudd, arnaddu, arnaddynt = arnynt; Pughe, in Coxes Monmouthshire, 1801, says: Gwentian


(delwedd B2712) (tudalen 131)


arnaddynt =Venedot. arnynt. This is confrmed by the medieval manuscripts, where, however, all these forms in -addunt are seldom used exclusively. In the new edition of the Mah. I counted 10 arnadunt, 16 arnunt, 11 attunt, but I do not guarantee these numbers to be absolutely exact. As to onadunt aud ohonunt, the different texts of the Ma). show very different gures: larlles y ffynn., 15 onadunt: 1 ohonunt; Ger. ab Urh., 25: 3; Triads, 9:1; Fer. ab Efr., 6:5; Br. Maxen Wl., 2: , Ll. and Llew., 6: , K. ac Olwen, 6: 7, but Fwyll Pend. B., : 6, Br. tercli Ll. : 4, Man. vah Ll. : 4, J/. ah Math. : 7. With this the relative proportions of a oruc and a wnaeth, etc, in the same texts should be compared. In the parts printed from Ll. Gw. Ehydd. onadunt largely prevails. In Hanes Gr. ah Cynan, Myv. Arch., cf. arnaddunt ISTorthern MS. arnynt, pp. 729, 731, attadunt wynteu attynt hwy, p. 731, etc.

[33.] Third sing. fem. arnei, arni,erni: cf. 5.o/Carm.,Skene, No. 13, erni {ih., imdeni, o heni, genti); U. (Gwentian Code), pp. 305, 365, 386, etc.; F.= Harl. 4353, f. 306; S. = 22,356, f. 5 (3); Owen.pp. 593, 600 (arneu, p. 558, -eu also for -ei of the pres. sec, pronounced -e); B. of Hcrg., col. 557, 635, 666, Skene, p. 251; Ll. Gw. Rhydd., pp. 135, 162, 167, 172 (2): arnei 164, arney 213; March. Crioydr. (ed, D. S. Evans), 17 cent, p. 257, etc.

Ernynt, like o henynt, occurs often in the later Gwentian texts, as Trioedd Dyfnioal Moelmud, Barddas, etc.; cf. Owen, p. 655 {Myv. Arch., ^929), Bardd., i, p. 64 (erni, Owen, p. 657). In Y Gwcithiwr (Aberdare): erni nhw, 1858, No. 7 (Seren Cymru: arni nw, i, pp. 233, 292).

[34.] Some other forms are: arnam, B. of Hcrg., col. 742 (2); Sal., N. T.: arnoddynt, f. 212a (see 16), arnaddwynt, f. 32&, 193&, arnynt, f. 212, arnwynt, f. 1736; arnywch, f. 376 (like wrthywch, gennywch); Addit. MS. 15,003, f. 2a, (Owen Jones): aruat, arnot. Arnef for arno ef is mentioned



(delwedd B2713) (tudalen 132)


by D. S. Evans {Lhjthyr); arno inaii, Y Tyw. ar G.; arna i, ardo fo, Yr Arw., 17, 7, 56. Arclo is an interesting form, of course quite niodern, vrhich I have not found elsewhere.

[35.] At: attadunt in Middle Welsh MSS.; etti, like erni, recdi: V. = Harl. 4353, f. 2b, Ua; Ll. Gw. Rhydd., pp. 141, 167, etc. Sal., N. T.: attwynt, f. 41; atwynt, f. 140&, 3626.

[36.] Tan: This preposition is nearly always compounded with others (am, ym, y, o, often a: a dan; also o-ddi- and inversely y a: y adanei, Ll. Gt. Rhydd., p. 24).

Cf. Addit. MS. 19,709, f. 9a: amdanadunt, f. 22a; ỳ danadunt (attanadunt, f. 29&, 41&); Cleop. B. 5: mdanaduut, f. 216; Ll. Gw. Rhydd., p. 35, etc.; ib.: amdanwynt, p. 227, the oldest example I know of -wynt; Sal., iV. T.: amdanwynt, f. 77; am danwynt, f. \nh; Y Drijch Chrst., 1585: amdanwynt; amdenyw, see Dosp. Ed., 939.

Ac adantaw ynteu, Ll Gw. Rhydd., p. 280. Amdeni is said in I)avies Gramm. to be a Powysian form; it is common in South Wales. Cf. imdeni, B. of Carm., Skene, No. 13; ymdeni, B. of Eerg., col. 826; deni, 843; ymdenei, 774 (ydanei, 600, 668, etc.); i^Ll. Gw. Rhydd., adanei, p. 7; y adanei, p. 24). Sal., N. T., amdenei, f. 186; amdeni, f. 3936; Cann. y C, 1672, deni, p. 204.

In the modem Dimetian dialect: am deni, Scren Cymru, ii, p. 505; plur, am deni nw, iii, pp. 206, 545 (cf. erni nhw); am dani nw, i, p. 272.

[37.] III. -yf. Can: e in cennyf, etc, is proper to all persons. This is due to the infecting terminations -yf, -yt, etc, as if its extension by analogy from the 3rd pers. sing. f em. (genthi ganthei, like erni arnei) were supposed, *ernof, etc, would be expected also; but they never occur.

Ganthu, 3rd pers. pL, B. of JLcry., Skene, p. 286; trỳchan meirch godrud a grỳssỳws ganthud, is printed (Skene, p. 99) from the B. of Aneurin, Gorchan Macldenu, in the handwriting of the second part of the Gododin. This ganthud


(delwedd B2714) (tudalen 133)


would be a form in -udd, the only one of whicli I ani aware. Cf. perhaps also trwyo and trwodd (= trwyodd), adv., through; the latter being a form equally obscure, if trwyo is to be compared with heibio and the explanation of heibio in Eevue Celt., vi (Rhŷs), is taken into account. Ganthudd, ganthu = ganthynt, Davies, Dict.

B. f Hcrg., Mah. (1887): genti, col. 556, 735, 745, 747, 754, 759, 823: genthi, 15 times; gantunt, 557, 569, 735, 737, 742 (2), 744, 810, 816 (gantaunt, 671, an error); ganthunt, 569, etc.; genny?i ni, 733 (with n due, as later in the verb, to the pronoun ni), gennwch, 733, gennch, 737, 785, 822, Cennwch corresponds to the 2nd pers. plur. of the verb (cerwch). It is not easy to find a link between the present teuse of the verb and the suffixed pronoun terminations. It can only be supposed that the coexistence of cerych in the pres. sec. and cerwch in the verb caused gennwch to be formed instead of gennych. Ll. Gw. Bhydd.: genti, p. 217; genthi, pp. 134, 144, 170, 221; genthy,pp. 251, 259 (crogy,p. 250, paratoy, p. 251, goleuny, p. 260).

Sal., N. T.: ganthwynt, f. 2236; a chanthwynt, f. 25; y canthwynt, f. 52&, 1756; centhynt,aty C.,ganwynt; genwchwi, f. 175; genwch, f. 269; genych chwi; cennwch,f. 358; cenychi, f. 348; genywch, f. 376. Gr. Eoberts, Gramm.: na chenthy- nhwythau, p. 68 (270); a chentho, p. (202); Athr. Crrist.: gentho, pref., gyntho, p. 51; Y Drych Christ.: gyniti, f. B 2, gantho, f. A 4, gyntho,f. A 2h, 21,etc.; genych, f. A 2; ganthynt, f. 196; gynthyu, f. A 26, etc. Addit. MS. 14,867: ganwyf fi William Morris, 1758; Addit. MS. 14,929: gennifi Lewis Morris, 1760 (see Cat. of the Addit. MSS., ii). Pughe and Dosp. EcL, 940: canof, canthof, cantof, canddof, canwyf, canthwyf, etc.

[38.] In the modern dialects gyda has partly replaced gann. Cf. L. Morris, Addit. MS. 14,923, f. 134: S. W. gyda = N. W. gantho. Also Dosp. Ecl, 1182: K W. y mae gennyf = Dimet. i mi (but i fi is elsewhere said to be S. W.)


(delwedd B2715) (tudalen 134)


= Gwent. gydafi (?). Eowlands, Gramm., gives N. W. y mae genyf lyfr = S. W. y mae gyda fi lyfr.

In Sercn Cymni (Dimet. dialect) gyda is nearly always used in popular language: gyda nw, i, p. 233; da from gd, the unaccented vowel being dropped; da fi, i, p. 373; ii, p. 364, da tithe, dafe, da ge, dag e, da ni, da chi, da nw; also da r hen bapyre newy ma, i, p. 372. Gwentian: geni, gen i, gento fe {Y Gwdtiiwr, Aberdare); genych chi, gentu nhw {Y ŵ, Llanelli).

In North Wales the unaccented form of gann, gynn, is used with the personal pronouns (gin ti, etc), as also rliwug fi, etc, for cennyf, etc Gin, the pronunciation of y in a monosyllable being expressed by i, is also transferred to the dissyllabic forms gennyf, etc: ginni(f ).

Cf. Addit. MS. 31,056, f. 137: mae ginine. Yr Arw. (Pwllheli): gin i, gyno fo, gyni hi, gynoch chi, gynun nhw, etc.; Y Gcncdl Gi/mrcig (Caernaryon): gini, gin i, a chyno fo, gyni hi, gynyn nhw 20, 5, 1885, p. 7, col. 6; gynon ni, 22, 4, 85, p. 6, col. 1. Sweet, Sp. N. W., p. 450: gini, ginti, gnofo, ginofo; gnoni, gnochi, gnynliw. Merionethsh.: cin i, gin ti, gyntho fo, gyuthon ui, -thoch chi, -thyn nhw, Cahan / evK Tomos. In this book also, p. 481, gynthwn occurs, about the only example I remeniber, due, of course, to carwn, cerwch of the pres.: gennwch.

[39.] Wrth. In p. 79 of my Beitr. I pointed out a number of forms in MSS. and books like gwenaeth, gwanaeth, for gwnaeth, etc. I admitted them only conditionally as historical orthographies, the pronunciation of gueneuthur and guneuthurin MS. . of the Yenedot. Laws being gneuthur, as written in the same MS. Or, and this seems to be the case in the 16th cent. forms (Salesbury, and Atlir. Grist.), they are the so-called inverse orthographies (umgekehrte Orthographie), brought about by the pronunciation of gwar- as gwr- and, gr-, etc I hold precisely the same views on some forms, gworth- gwyrth, occurring mostly in later texts, for gwrth, wrth. They are etymologically correct, as gwrth springs from *gwerth, *vert-; cf Corn. worth, etc.


(delwedd B2716) (tudalen 135)


Cf. LI. Gw. Rhijdd., p. 236: rac dryccet y bobyl a wyrthwyneppyiit iidunt (perhaps a scribal error caused by the following wy); Addit. MS. 14,973 (1640), Rees Prichards Pocms, . 67h: gweddy worth fyned yr gwely; Add. MS. 14,974, f. 6o: i worthladd drwg; Foxs Battledoor, 1660: a giliasocli oddywurth iaith ych mame, p. 6 (Gwent. dialect; firiadau, ihiddu=beiddu, hwunt, gwibod, etc).

[40.] B. of Herg.: rtliywch, col. 784; rthych, 727; Tit. D. 22, f. 156a: a minneu a ymcholyaf (leg. ymchoylaf) vy wyneb y rthych a y 6rth ach tei a rywnaeth ych d6y hi6; f. 158: kanys mi a dileaf bob dr6c a goueileint y 6rthy6ch. Sal., N. T.: o ywrthynt, wrthynt, wrthiut, f. 6, wrthwynt, f. 178a.

In modern dialects: wrthw I, Sercn Cymru, ii, pp. 6, 262, 382; wrtho i,ii, p. 382; wrtho ti, ii, p. 382; wrtho ni, i, p. 232; wrtho chi, i, p. 272; wrtho ti, etc, Y Gweithmr (Aberdare); wrthw I, Y Tyw. ar G., ii, p. )); wrtho i, i, p. 96; etc. In the Carnarvonshire dialect wrthaf is used: wrtha i, Yr. rw.{vf\\- heli), 17, 7, 56, wrtha ti, o wtha ti (cf. wth welad, wth y droull, wth y mhen i, wth i gilidd, wth ddwad, etc, ih. The omission or indistinct pronunciation of r is frequent, mostly in unaccented syllables, proclitics, etc, in the colloquial language), wrtha chi and wrtho chi, etc. In Cahan few. Tomos: wrtha i, p. 7, wrtha ti, p, 56; wrthat ti, wrtha titha, etc Addit. MS. 31,060, f. llh (N. W.): wrthat i. Wrthaf may be explained like wrthof, etc.; but as the pres. sec in the Yenedot. dialect ends in the plur. in -am -ach -ant (froni -em -ech -ent, the alternating forms of -ym, -ycli, -ynt), the -af- tlection in this dialect can also be explaiued by wrthym, etc, foUowing the example of the pres. sec.

[41.] IV. Oc: This preposition is not immediately compounded with the pronominal terminations, but the form ohou- is used before them, which has, with regard to its nasal, a parallel only in arn-af, if this is to be compared. In Zeuss, ^p. 666, ohon- is said to represent an older *ocson, which is possible (cf deheu); but the further comparison of Ir. , as


(delwedd B2717) (tudalen 136)


(l. c.) is wrong. Ir. , ass, is the unaccented (proclitic) form of the preposition es-, Welsh, eh- (Exomnus, ehofyn, esamain); cf. Zimmer, Rcische Studien, ii, p. 91. Ohon- is pre-Cymric; cf. Corn. and Bret. ahan- (preposition a, Vann. ag). I think c (from) is identical with oc, ac (with), Ir. oc, ac, the intermediate steps, as regards the meaning of the preposition, being: with (sociat.) by (instrum.) from (without regard to an instrumental relation, denoting simply the origin). Son- I . always thought to be of pronominal origin, connected with Welsh hwn, but I cannot prove it. Ohan- is very common, besides ohon-, and is due to the same causes as not a few other instances of a from o. Also ahan-, like the Corn. and Bret. forms, is met with.

[42.] I have not found a single example of ohan- in tlie Oxford Mahinogion, and in the texts printed from Ll. Gw. Bliydd. Other texts I have not examiued strictly with regard to this question. Ohan- occurs in a manuscript as early as the BooJc of Carm.; cf. o hanaut te, Skene, No. 18. A few other instances of olion- and ohan- are: B. of Tal., o honat, Skene, No. 2; B. of Herg., o honat titheu, Skene, p. 177; o hanei, o honom ni, p. 179; o honach ch6i, p. 195; o honofi, col. 622; B. = Tit. D. 2 ( Venedot. Code), o hanav, f. 1; o hanaf u, f. 705; o hanaf uỳnheu, f. 66 (ohan- mostly used); U., ohoni, p. 360; V., o honei, f. 33.; ohon- mostly used in Addit. MS. 19,709, Clcoj}. B. 5; Tit. D. 22 (Dimet.): o hanot

ti, f. 170&; o hanam, o hanam ni, f. 1476.

In later texts: Sal., JY. T.: o hanof vi, f. 16; o hanaf, f. 16; o hanat, f. 74&; o hanot, ff. 5h, 11&; o hanam, f. 217. In the parts of the N. Test. translated by E. Davies ohon- occurs oftener than in the others; Huet: o hi (hanei), f. 393; o honi, f. 391; o hanoch chwi. In the 16th cent. S. W. Medic. MS., Addit. MS. 14,91 3: yr hwnn arverir o hana (sic) yn erbyn, f. 1. In other handwriting: o hanynt, f. 11; a bwyttaet gwrs a lianaw bob borc, f. 276 (o hanaw, . 27, 29, etc); cf.


(delwedd B2718) (tudalen 137)


Ehŷs, B. Celt., vi, p. 45. In parts of Soutli Wales aliaiia -i (Ist pers.) occurs.

[43.] As to -af and -of in tbe Mabinogion (1887), -of is extremely rare (o honof i, col. 836), whilst at least 10 -af and 11 -at occur. In the Ll. Gw. Rhydderch, -of is of much more frequent occurrence; cf. 7 -af, 8 -at, 8 -ot, 9 -am, 2 -oni, etc.

[44.] On onaddunt, see 16, 32. The Corn. anozo, 3rd sing. raasc, anezy fem., aneze 3rd plur., aud the Bret. anzhan, anzhi, ane^h (Lon.), anehou, anehi, anehai (Vann.) show that -on from ohon- is pre-Cymric. I think it is due to the accent; *ohon-o-t (see 16) lost the second o, hence *o(h)notu, onaddu (dd not explained). Can n in arn be explaiued in the same way? (*are-son-, *ar-hon-j *ar(h)- nad?). There occur in 8. (Addit. MS. 22,356), Cleop. B. 5, and Sal., N. T., some apparently open forms, o anadynt, o honadynt. As they are found in prose texts, I am not able to decide whether they are etymological orthographies or whether olion- has really been introduced into onaddunt from the other persons.

Cf. o anadynt, S., f. 76a; o lionadunt, Cleop. B. 5, f. 15 (2); o hanaddynt, Sal., A^. T., f. 62, 88, 1445, 253a; o hanoddynt, f. 152a; o hanaddwynt, ff. 189, 28a; o naddynt, f. 71a; o hanyut, f. 195a; a hauynt, f. 385; o hanwynt, ff. 32a, 535, 885. Ouadduut occurs like the third pers. sing. pret. iu -ws, said to be S. W., also iu the Venedot. MS. A. of the Howel. Laws: o nahunt, p. 56; ouna- dun hui, p. GO; and onadH<, pp. 63, 73, 75 (2), 77. Ou this -uuut I cannot rely, as it uiay have been misread for -uunt.

[45.] In the 16th cent. onof, onot, onofo, onom, onoch, onynt, are frequently used in Griffith Eoberts books. I do not remember to have found them any where else. He uses also honof for ohonof; so onof is from honof, and has, of course, nothing to do with the old onaddunt. Cf. Athr. Grist.: ar wneuthur honomi ewllys duw, p. 26; onofi, p. 5; onof o, p. 7; onoch, p. 7; onynt, p. 36; Graiiim., rhai honynt, p. (186);


(delwedd B2719) (tudalen 138)


onof i, p. 35; onoti, p. (161); onaw, p. (183); onofe, p. (95); ononi, onyn, onynhwy, p. 52; onyntwy, p. 121, etc. At tlie same tinie monof (from ddimhonof, (dd)monof) makes its first appearance (in Gr. Eoberts books). It is now frequently used in the N. W. literary language of no puristic aims. In the spoken language also, nio for ddim o is used; cf Gr. Eoberts, Gramm., p. (389): moth henaint; Hughes, An Essay on the .... Welsh Language, 1822, gives mo (na ddywed nio hynny) as N. W.; Sweet, Spohcn N. W., p. 442: welisi mm td, m d dd, etc.; na . . . . monun nhw, Yr Arw., 17, 7, 56. SaL, N. T., does not use monof; he very often has ddim hanof: na ddanvonei ef ddim hanynt allan or wlat, f. 50&; nid adwaen i ddim hanaw, f. 124a; ni wrandawawdd y deveit cldim hanwynt, f. 149 (pwy n hanoch, f. 147a). I have never met with *manof, and I think ddim hanof is S. W.; cf. Lewis Morris in Addit. MS. 14,923, f. 133: S. W. ni bu ddim o hano = N. W. ni bum i; also in Y Traethodydd, iii, p. 13: S. W. nid oes dim o hano i yn ei nabod (= nid wyf i yn, etc.); nid oes dim o hano fe yno ynawr; cf. Seren CymTu, i, p. 373: weles I ddim o nhad yn ddarllen un llyfyr eriod.

[46.] heni, B. of Carm., Skene, No. 13, 18; U., p. 360 (Owen): o honi =W.o heni (Cleop. A. 14, e.g., f. 9Sh); Q., pp, 556, 559: o honei = P., P. and S., S. o heni (ib. Q. arnei F. erni); S., p. 602; V. = Harl. 4353: gedy yd el yr eil heit o honei, and gedy ydel y tryded heit o heni, f 33a(erni, f. 30?*); B. of Herg., col. 737: o heni hitheu, o henynt (cf. ernynt, 33); Owen, pp. 639, 676 (o heni, pp. 662, 667); Bardd., , pp. 24, 32, 102, etc; Llyfr Achau (Breconsh., 1602), p. 17; Scrcn Cymru: o heni nw, iii, pp. 207, 266, 465, 544.

[47.] OnadUj B. of Tal., Skene, p. 51; B. of Herg., Skene, p. 264 {ih., racdu, col. 735);, W. = Cleop. A. 14: vn o honu o talu or taladỳr drosta, f 74a; ih., pop vn o honu, Owen, p. 309.


(delwedd B2720) (tudalen 139)


[48.] In modern dialects: oedd llawer o nw n meddwi, Monmouthsh. (preposition o); o hono ihunan. o honon ni, o hono nina, o honun nhw, Yr rw.; o hona i, o honoch chi, o honyn nhwthe, Gab.few. T.; Y Geninen, i, p. 160: o honwyf (literary language).

[49.] V. Do: udu, udunt (3rd person phir.) are the regular Middle Welsh forms. I cannot help assuming that y in *ydu (do-) has been assimilated to the second li (udu), as a genuine u in udu cannot be explained in any way. In later Welsh, y and i, from the 3rd person sing., are reinstalled in the plur, (yddunt, iddunt).

Cf. udu, B. of Ta., Sk., Nos. 51, 52; B. of Herg., Sk., p. 265; uddudd = iddynt, interdum vddu, Davies, DicL; Cleop. B. 5: vdunt; Tit. D. 22: vthvnt6y, f. 166; S.: udynt, p. 592; ydynt, p. 594. Sal., N. T.: yddyn (ofteu), yddwynt, f.olGh; yddwynt wy, f. 225a; vddynt, vddyntwy, etc; Gr. Roberts, Athr. Grist.: megis y darfu iddynhwythau, p. 47; J. D. Rhŷs, Gramm.: iddint, yddynt, uddunt (dialect.); 3rd pers. sing. fem., iddei, idd, p. 128.

In modern dialects: iddi nw, Ser. C.; iddinhw, iddiut (Aberdare); iddun nhw, iddu uhw, Yr Arw.; g ithi hi, Cah. f ew. 7., p. 200.

[50.] In Y TraetJiodydd, 1869, p. 20, ymy, yty, yny, are quoted from E. Pryss Fsalms (Meriouethsh., first edit. in the Bible of 1630); the forms now used are, it is said, imi, iti, ini, and more emphatically, i mi, i ti, i ni. Perhaps these older forms are preseiwed in ethi, ethyn (cf. efo ag ethi hi, gydag ethyn nhw, with c for = y), cited in YTraeth., iii (see 4), from N. W. dialects. L. Morris, Addit. MS. 14,923, f. 133, has S. W. i fi = N. W. i mi; the same is given in Y Tracth., iii, p. 13, and 1870, p. 412 (i fi, in WilHams Pant y Celyns Hymns). Cf. also Scrcn Cymru, i, p. 232, and i maes, i faes, i mewn, i fewn, etc.

[ol.] VI. Eidof: Eiddof, einof, etc, seem to have taken their origin from the 3rd pers. sing. eiddo, a form very difficult of explanation. If y meu fi, the mine, is tlie mine to me, fi


(delwedd B2721) (tudalen 140)

140 OBSEKVTIONS on the welsh peonouns.

being in a dative relation to the poss. pron., eiddo is perhaps ei iddo, his to him. At any rate the possessive ei was believed to form a part of this form, as the poss. Ist pers. phir. ein was transferred into this forni with the suffxed pronoun; einym, einwch, are, I think, much later forms than eiddo, and are formed after its analogy. I do not attach any value to ei ddaw in SaL, iV. T., as he has many merely etymological orthographies; cf. at yr ei ddaw y hun, f. 131&, ar ei-ddaw yhun, etc. Salesbury also uses yddo, caused by the coexistence of the poss. prou. ei and y. Cf. canys yddo duw ydynt (duw ei pieu), f. 249; ar yr yddoch ychunain, f. 294a; nyd ydywch yddoch ych hunain, f. 249 (yr petheu eiddom, f. 325&; eiddwynt, f. 6b).

Davies, Did., has eiddo, goods,bona, orum; L. Morris, Addit. MS. 14,944, f. 795 {Additions to Davies Dictionary), says it is a Dimetian word. It does not occur in the Carnarvonshire dialect; cf. Sweet, p. 442. In Spurrells Dict., also, the infin. eiddio and adj. eiddiog occur. Eiddof, einof, einwyf, etc, are given in the grammars.

[52.] There seems to have been a time when eiddo (Middle W. eidaw) was not very strongly connected in the mind of people speaking, or at least writing, with the preposi tion with suffixed pronoun, as idyaw occurs in almost every Middle Welsh manuscript.

Cf. A. (Venedot. Code): Eammereth paup er eidiau val kent, p. 53; eidya TJ., pp. 303, 330; ae eidia yuteu, *S., f. 4?;, Owen, pp. 595, 613; B. of Herg., col. 589: yny eidia eliun; Ll. Gw. Ehydd.: eidyaw, p. 57; Y Seint Greal: yr eidyaw ehun, p. 411; yn yr eidiaw, p. 196; castell a vu eidyaw dy dat di, p. 300; or bei gyn chwannocket ef yr meu ac ydiw gennyf I yr eidyaw (priuted eidywaw) efo, p. 431; Cleop. B. 5: eidiaw, f. lb, etc.

I am not certain whether y (j ) was really spohen here, as it was inserted indiscriminately by the South Welsh scribes, who did not pronounce it (-yon: -on, -yaw: -aw, etc), and who therefore lost the proper feeling of where it was justified and


(delwedd B2722) (tudalen 141)


where not. So in Oderics travels, in Didrefn Gasjliad, dwylyaw occurs several times (f. 122&, pp. 244, 245, 252; d6y- la, p. 266); biit I never fouud a preposition with suffixed pronouu inserting y (j), except trydya, Bidr. GasgL, p. 245.

C. The Possessive Peonouns.

[53.] Ist sing. masc, my, fy, y, before consonants whicli have been nasalised, occurs very often since the sixteenth century. In the colloquial language fy or y are of ten omitted altogether, the nasalisation of the consonant alone indicating the poss. pron.

Cf. Gr. Roberts, Gramm., p. 70 (302): mae n dda y myd and mae n dda myd for fy myd (byd); Addit. MS. 14,986 (16th cent.): gair ymron, f. 12; Cann. y Cymry: 1672, p. iOO: a gwella muched; Cab.f ew. T.: ja Y nghalon, p. 66, etc.

[54.] The forms of the poss. pron. Ist. pers. sing. before Yowels are, according to Ehŷs, LecL, ^p. 52: N. W. / (f enw), S. W. fyn (fyn enw), in North Cardigaushire, fyng enw.

Cf. Addit. MS. 15,059, f. 2236 (1760): fy nwylyd fenaid, fy nwyl(y)d fenaid i (ib., mynad, cerddad); IV Arw.: ma nhewyth. Seren Cymru: newyrth, iii, p. 142; yn cal n ffordd, iii. p. 5; a newrth (Aberdare).

[55.] The following sentences quoted from modern dialects show the very curious nasal inflection of verbs: Oahanfeyj. Tomos: waeth gin i pwy nghlowo i n i ddeyd o, p. 61; ac ar liyn mi nhrawodd i yn y mhen, p. 61; waeth gin i pwy nghlowo i n deyd hyny, p 155. Seren Cymru: ac am ddeg yn y nos mi nhowlson i mas, iii, p. 524; also Scren Gomer, 36, p. 584: A fi ngweles I weithian yn myndgyda lot o fechgyn. Here mi, fy (nasalising the verb) + verb -I- i (fi), seem to be formed after the analogy of ei -i- {uninfected) verb + ef; cf., e.g., from Sal., iV. T.: ac wynt ei gwatworesont ef, f. 14a; a hi ei duc yw mam (yddy), f. 23a; ac ei cernodiesont (bonclu- stiesont), f. 45, etc.; pan ei gwelwyd = pan welwyd ef,


(delwedd B2723) (tudalen 142)


etc. I liave never found tliese syntactic anomalies in otlier texts.

[56.] Meu, teu, are not used now with substantives. This usage is frequently met with in SaL, JS. Test.; cf. y tuy meu vi, y gairie meuvi ar law veu vi (am llaw vy hun), f. 123; cf. also Edm. Pryss Fsalms (according to Y Traeth., 1869, p. 205): fy anwiredd mau, y min mau ( = f y min i), yr oes fau; yr ebyrth tau (= dy aberthau di); dy foliant tau; llawenydd hwn mau fi a gyflanwyd, etc.

[57.] The possessive pronouns, Ist and 2nd pers. plur., are an, yn, ein; ach ych, awch, eich. They all occur in Middle Welsh texts; cf. o nerth an brecheu ac yn kledyfeu, B. of Herg., col. 620; o waet an hargiyd ni, col. 607; yn ach clusteu, Ll. Gio., p. 260; yn ach plith chwi, Tit. D. 22, f. 1555; ar ych llafur, f. 1576; y dodi ach kyrff . . ., f. 156, etc. A close investigation of their syntactic differences and relative age is much wanted. An and yn seem to be accented and ■unaccented forms like ei and y, preposition am- and ym-, etc. Ein, eich, are transformations of an, awch, after the mitster of ei (3rd pers. sing.); the Ist and 2nd pers. plur. seera also to have followed their own mutual analogy.

[58.] Ei and y have been explained as being the accented and the unaccented forms (Kuhns Zeitschrift, xxviii). In Llyfr Gweddi Gyffredin, 1586, pref. (see my Beitr., p. 24, 3), S. W. i, y, is noticed. Davies, Gramm., p. 179 (ed. 1809), gives the prouunciation i of ei; also Rhŷs, Eevue Celt., vi. The connection of this poss. pron. with y (do) is of much interest, as there exists a startling variety of forms, which are not all noticed in Zeuss^. I know oe, oe y, iw, iw y (iw ei), iddei_

[59.] I. Gc (to his, etc). This is the regular Middle Welsh form, used almost without exception in tlie Ma)., Ll. Gio. Rhydd., etc. Some examples are: B. of Carm., Skene, No. 18: inỳdel kinon iti oe chingueled; Tal., Skene, 33: k6r6f oe yfet; Addit. MS. 19,709, f. 10&: ac yn menegi hỳnỳ oe


(delwedd B2724) (tudalen 143)


gedymcleithon; T. Gw. Rliydd.: j jmgeissaw ac ef ac oy rwymaw, p. 78; oy claly y dauaw, p. 81; oy gwrandaw ef, p. 84; na deuey ef byth oe gyuLawn lewenyd, p. 94; ae doyn oe grogy, p. 256; Tit. D. 22: kanys du6 ehun a anuones yr ysgriuenedic rybud honn yr pechaduryeit hyt ar allar eglys pedyr a phal in ruuein oe rybudya amweith sul a gyl, f. 15Sh, etc. Oe is also used in the 16th century, but I am not aware of later examples; cf. AtJir. Grist.: beth sydd oi obeithio gentho, pref.; Y Drycli Ohrist.: rhoi bustl idho oi yfed yn le gwin, p. 3; mal y mae llawer or lyfre hynn etto in plith ni oi gweled, f. C 1; Addit. MS. 14,913, 16th cent.: oe yuet; Addit. MS. 14,986, f. 14&: ni ai rwymwn ef oi eiste; 31,056, f. 30&: a gown a march oi wasnaethwr (Araith Gwgan). Gr. Eoberts, who frequently uses oe, oi, says, in his Grammar p. (129): iw is u.sed in sir y phlint (ef aeth iw wlad), oi in Carnarvonshire (e dymchwelodd oi wlad, rediit in suam patriam). J. D. Ebŷs, Gramm., 1592, gives iw, yw, oe garu (p. 7); idh ei, oi, iw garu (p. 125).

[60.] II. Oc y: L (Dimet. Code), p. 200: ac ny ellir eu hyraell o neb fford ytalu nac y erbynnyaw (gymryt, /.) dim dros (o, /.) alanas = 22,356, f. 1h (ac ny ellir eu lcymell o n. ff.) y dala dim o alanas nac oe y gymryd; Ll. Gw. Rhydd. {Bown TTctmtwn): ac y gofynnawd hitheu idaw beth a dybyei ef am y palmer . . . . ac erchi idaw mynet oe y edrych, p. 142. These very rare forms are supported by the existence of iw y, iw ei; see below. Oc ei becomes also oe (written once oe y in Cleojj. B. 5: o doeth anghev yw vrawt oe weithret ef, f. 129 = 3yv. Arch., ^p. 663&). There are in Ll. Gw. Bhydd. some very curious examples of oc ei used in the sense of to his; cf ac anuon dy archegylyon lcyssegredic yn eu kylch; oc eu diffryt rac tywyllwch. ac eu dwyn y deyrnas nef, p. 110 (to defend them); a gwedy hynny yr neuad yd aethant oc eu bwyt (to their meat), p. 152 (cf. ac yna yd aethant gwyr ffreinc o eu gwlat, p. 18; ac oe bwyt


(delwedd B2725) (tudalen 144)


yd aethant, p. 1 54); y sawl ny byryawd neit yn y mor oc eu bodi, p. 116 (he brained . . . those who did not leap into the sea to be drowned). If these forms were to be relied iipon, all that has been said and that wiU be hereafter put forward on the subject of oe, would be a failure ah initio. But I cannot make the sense of to his and from his agree; and so I think that, oe being the common result of both do ei and oc ei, some mistakes were occasionally made by transcribers, who wanted to use oc for o, and did so at times in the wrong place.

[61.] III. rio. Yw, iw, is now the only form of the literary language, except in South-Welsh books, where iddei may be found. Examples are needless.

[62.] IV. Iiu y, iw ei. The nearly constant use of yw y forms the chief peculiarity of the language of the Dares PhrygiiLS in Cott., Cleop. B. 5. This text is not connected with the other parts of the MS., on which see my Bciir., pp. 13, 14, 16.

Cf.: pantus a ynegia y priaf ac a danllewychws yw y gyfuesseiuyeit y petheu a glywssei ef gan euforbus y dat, f. 225a 1; priaf a rodes kennat yw lu (y before lu is erased: the upper part of it may still be discerned), f. 2256 1; priaf a anioges yw y eibyon, f. 2236 1; y gribdeilyaw elen ac yw y dwyn ganthunt, f 2266 1; yw y logeu, f. 22661; ac a erchis yw y gedymdeitiiyon, f. 2306 1; ac a erchis dyuynwu antenwor yw y (sc. poluxena) cheissyaw ac yw y dwyn attaw, f . 249a 1; and in eight other places.

In other places y eu is used: ac a ffoassant y eu hestyll gwyr troya, f. 2o9a 2; a gwedy dyuot gwyr groec y eu lluesteu, f. 239a 2, etc. The only other manuscript in which I found such a form is Addit. ]MS. 19,709 {Dares Phryfjius, Bed BooTc version; see my Beitr., p. 14): Ac ynteu a erchis y6 y 6ab a oed yn sefyll gyt a hi gal6 ector drae y gefyn, f. 4& = B. of Rerg., coL 17: ac ynteu a erchis y vab a oed ynsefyll gyt ahi gal6 eetor draegefyn.

[63.] Such forms are not wantiug in the modern dialects.


(delwedd B2726) (tudalen 145)


lu tlie passage quoted above, 14, from Tnjsorfa Chnyhodaeth, 1770, iw eu gilydd, iw eu dwylaw, are mentioned. In Cadioedigaeth yr laith Gijmraeg, Bala, 1808 (according to the Brit. Mus. Catalogue by Dr. Owen Pughe), on p. 26 are given: poss. ei, Yenedot. i; iw, iddei, ac arferid wy yn fynych gynt, h. e. iwy. Also in Y Traeth., iii, p. 14, and in D. S. Evans, Llythyriaeth, iw ei, iw eu, are given as forms to be avoided !

[64.] V. Iddy, iddei iddeu. These forms are said to be peculiar to the Gwentian dialect; they are of frequent occurrence in Gwentian texts since the 16th century. Of older examples, if Liber Landav. be excepted (on which see below), I am only aware of a few in the Dimetian MS., Tit. D. 22; cf. pobun yth ygilyth, f. 9a; ac yny rodi yn hoU iach id y vam (B. Dewi S.), f. 150. In Sal., N. T, yddy is often given in the margin, but in Huets Giueledigaeth Icuan it is ordinarily iu the text.

Cf. y ew duy ehun (yddy), f. 13a; a hi ei ducy w mam (yddy), f. 23a; yw (yddy) vrawt, f. 289, etc. Huet: yw ddangos yddy wasnaethwyr, f. 373; yddy wasnaethwr, yddy (yw) lle, f. 387a; yddy (yw) wasnaethwyr y proff wydi, f. 384;, etc. Y Drijdi Christ., 1585: ny bythont yn perthyn idd eu gwlad nhwy (printed nhyw); ynhwy a roesont . . . y groes idhy dwyn ir . . . lle, f. 31a; a roesont idho idh ei yfed, f. 315, etc. Addit. MS. 14,921, 16th cent. fJohn Maundeinlles Travels), Gwentian dialect (see my Beitr., p. 83): yddy chyssany, f. 4b; yddy fwytta ac yddy loski, f. lOa; yddy gweddie, f. lOa; yddy bwyta, f. 11; yddy ben y hyn, f. 14?); yddy chwer, f. 23a; yddy ddistiblon, f. 24a, etc. (yw gladdy, f. dO; yw pene, yw gilidd, etc). Medcl. Mijddjai: a dyro r claf iddei yfed, p. 92; dyro r plentyn iddei yfed, pp. 99, 100, etc. Bardd., iddei galw hi, i, p. 10; pp. 14, 54, 94, 122, etc. March. Crwydr. (ed. D. S. Evans): idd eu delw hwynteu, p. 7; idd ei fam, p. 150; yn nessaf ydd y croen hoeth, p. 257 (= y r, see below; hoeth for noeth is S.W.). Cann ij C., 1672: idd i weison, p. 121; idd i eglwys, p. 291; idd eu plant, p. 294; St. Hughes: idd i bwytta, etc.

J. D. Rhys, Gramni., see 59; Davies, Gramm., p. 173 (ed. 1809) iw = Dimet. idd ei, eu (also Richards^GraHw., 1753, p. 58); Pughe,



(delwedd B2727) (tudalen 146)


in Coxes Monmouthsh., 1801: Gwent. iddei ben =Veuedot. i w ben; Hughes, 1822: S.W. aeth iddei dy; Dosp. Edeyrn, 1260: Gwent. and Diraetian iddei; in the dialectal list, however, it is only said to be Gwentian. Cf. also Y Traeth., iii, p. 14.

In modern dialects: iddi fam, iddi mham (Aberdare; mhisS.W.). Iddei, iddeu is always used by lolo Morganwg in Y Cymmr., iv, pp. 101-5 (Glamorgansh.); beth sydd genych chi iddu wed (= iw ddywedyd) yn awr (Llaelli); idd i hercyd nhw shatre (Y Bedyddiwr, viii, p. 108, from ISIonmouthsh.; shatre = tu ag atref; hercyd is S,W.), etc.

An accurate delimitation of the parts of Soiitli Wales in which iddei is used, would be of great interest for the locahsation of many manuscripts, etc.

[65.] Iddei for iw caused, in the South Welsh modern language, i with the article (ir) to be supplantd by idd y, idd- being abstracted from idd-ei. In Traetiawd ar laum-

lythyreniad yr iaith Gymmracg, gan John Jones

(Ehydychen, 1830, 8), idd and odd are said to be regularly used in the S. W. vulgar language before the article: awn idd y tŷ, for ir ty or i y ty (this latter form is only a construction of the author). D. S. Evans, Llythyraeth, mentions odd ei, odd eich, odd eu, for oi, och, ou; and iddei, iddeu, iddein, iddeich, iddy, for vf, in, ich, ir. Eowlands, Gramm., ^. 118: ydd y tŷ, odd eu tai. These forms really occur in the literary language of South Welsh periodicals: cf. Seren Gomer (Swansea, 1830), p. 80: gyda phob parch idd eich gohebydd; xxiii, p. iv (written by the editor): idd ein gohebwyr; vol. v, p. 364 (1822): neu cyfieithiad W. S. ac y Dr. M. odd y Bibl Cymreig; ib. Dr. Pughe is said to write always idd yr; vol. xxvi, p. 271 (1843): some write am beth ag oedd yn bodoli, for am b. oedd yn b., odd y tŷ, odd ei ben, idd y dwfr, etc.; Y Bedyddiwr (Cardiff), 1851, p. 10: ac a adroddodd yr achos oddei ofwyad. These forms are not difficult of explanation. Iddei, iddeu, was separated into idd- and the poss. ei, hence idd with other possess. pronouns and with the article (idd ein, eich, y). Oddei, of course, follows iddei, as the full form of


(delwedd B2728) (tudalen 147)


the preposition is oc. Odd- in oddei was eitlier wrongly abstracted from oddiwrth, oddiallan, etc. (: iwrth), or and this is more probable oc ei was transformed into odd ei after the model of idd ei, by reason of the meanings of the two prepositions (to and from) being strictly opposed to each other. Such contrast of meaning very often contributes to the mutual assimilation of two words or forms; a large number of examples of this kind of analogy were collected by Brugmann in his article on v, iv, et (Berichte der Schs. Ges. der Wiss., 1883, pp. 181-195). Irish examples are ssar sinser, tess taid, etc. (see Stokes in Bezzenbergers Beitr., ix, p. 92); cf. Welsh asswy and asseu deheu.

[66.] In Lihcr Landavensis some instances of iddy seem to occur. This is not surprising, as the probable explanation of iddy, iddei, requires the assumption of the existence of the initial d of the preposition, which is indeed sometimes kept in L. Landav., a Gwentian text. I shoidd put fuU. confdence iu these forms, if this text were more carefuUy edited and from the original manuscript, still in existence. In the great charter on pp. 113, 114: ac idythir hac idi dair (of eccluys Teliau). In the same charter: har-mefyl har sarhayt, etc, a guneel brennhin Morcanhuc hay gur hay guas (and his ) dy escop Teliau hac dy gur (to his ) hac dy guas (to his ); ha diguadef braut diam y cam a diconher dy escop Teliau ha dy (to his) gur ha dy guas. If these dy, to his, are to be relied upon, they are forms of very great importance, and, as far as I know, unique.

[67.] So we find oe (oi), iw, oe y, iw y (iw ei), y ei, eu, iddy (iddei), and perhaps dy, used to express to his, her, their. In the foUowing paragi-aph I will outline some guesses on the connection of these apparently disparate forms. The chief difculty is the form of tlie poss. pronoun. *Do- becomes *di (cf. Corn. dy, Bret. de, and probably Welsh dy in L. Landav.). To this, at the time when the initial d



(delwedd B2729) (tudalen 148)


stiU existed, at least in sorne positions (sandhi), as in L. Land., di was prefixed, giving di-dy, i-dy, later iddy, and, later, iu the accented form of the poss. pronoun, iddei, iddeu. Cf. Ir. chucmn, chucutj etc.? D-ei, do-i, gave oei, oe, oi, which existed till the 16th century, if not later. The reasou of its disappearance is probably the exclusive use of oi, oe, for from his, after oc (oc ei) fell out of use in North Wales. In South Wales iddei prevailed, and oddei for oc ei followed its analogy. The explanation of iw offers by far the greatest difculties. Rhŷs, Revue Celt., vi, p. 57 c seq., explained bwy in bwy gilydd, etc. (or mor bwy gilydd, etc, often in Middle Welsh texts), from po-i (i being the poss. pronoun). The colloquial form is bw. Inthe Book of Tal., Skene, pp. 138, 154: y ren ry digonsei, r6y digones; r6y goreu, p. 158; n6y kymr6y, p. 147; rwy golles, Myv. rcJi., ^p. 160a {Cynddelw), etc; cf. Ehŷs, Revue Celt., vi, p. 50 et seq.: rwy-, nwy-, are rho-, no-, and the infixed poss. pronoun i. The conditions of accent, etc, under which in pwy, rwy-, nwy- wy, sprang from *oi are not known to me; but these examples enable me to assume *do-i (tohis) becoming under certain circumstances *dwy, *wy {d lost). There exists ry6goreu, B. of Herg., Skene, p. 233, confirmed by Rhŷs, l. c. If this form is to be trusted, and if others similar to it exist, it would be possible to assume yw (written yw) to come from *wy (do-i). But I prefer the following explanation: the loss of the initial d is certainly due to the influence of the different final sounds of previous words and their greater or lesser syntactic connection with *do. If, therefore, di exists in so late a MS. as L. Landav., this is no argument against assuming a much earlier loss of d in this preposition in certain positions. To this *wy, as in i-dy *dy, y was prefixed, giving *dy\vy, *yẁy. The usual forms, yw, iw, are to be explained Like bw from bwy, -ws (3rd pers. sing. of the s. pret.) from -wys. By the way, I think this South Welsh -wys to be formed after the analogy


(delwedd B2730) (tudalen 149)


of -wyt (part. pret. pass.), since -as -es -is, and -at -et -it (-wyt) coincided in vowel. -Wyt itself is not clear at all; it seems to be a wrong abstraction from ttie part. pret. pass. of the verb substantive, containing, besides the sufx, a part of the stem of the verb. The question now rises as to oe y, yw y, and iw ei. It is a very seductive supposition, to take y w y in MS. Cleo]}. B. 5 (see 62), for the y-wy which I thought to be the source of yw, like -wys, so often written besides -ws. Is the modern iw, ei, eu the successor of an older iw-y, like iddei, iddeu, and the older iddy? Or was i^v too obscure, and ei, eu reintroduced to enforce the possessive meaning? And is this perhaps also the case with yw y in Gleop. B. 5? Oe y almost points in this direction, if it is to be relied upon at all, of which I am not whoUy convinced. I am not able to decide this question, but this does not tend to disprove the other assumptions. In conclusiou, I would draw attention to py (or mor by gilydd), occurring besides pwy (bwy, bw); bwy (bw): by, correspond exactly to *wy (yw y, yV): dy .(Z. Land., Corn., Bret.; i-dy in South Wales).

As to the explanation of iw by Rhŷs, Bev. Celt, vi, if it is to be preferred to that proposed above, I would rather explain the V in the poss. 3rd pers. sing. by an analogical transfer from the teu (*tevos) of the 2nd pers. sing., as was the case with meu (Ist sing.), and Ir. mo (see Kuhns Zeitschrift, xxvii, p. 401, note 1).

D. ThE PtELATIYE PrONONS. [C8.] The relative pronoun a in its double function as simple relative and as so-called verbal particle (in construc tions where its original demonstrative sense has faded away see Zimmer, Kelt. Studien, ii, p. 59) is supplanted by y in South Welsh, niore especially Gwentian texts, sincethe 16th


(delwedd B2731) (tudalen 150)


century at least, I am not aware of any earlier instances. As to the explanation, see perliaps Eliŷs, Ledures, ^p. 147. Y, saicl to be the form of the oblique cases of a, has exceeded its proper domain of use; the reasons and the history of this analooical transrression are obscure to me.

Cf. Lljfr Gweddi Gyffredin, 1586, pref.: a being relative or a Yoyce expletive for y, not used to them of South Wales; and y being rel. or a kind of exp]. for a, vnto the North Welsh readers. J. D. Rhys, Gramm.: mi a garabh, peth a gaffer, and mi y g., p. y g. Davies, Gramm., p. 182 (1809): Dimetian mi y garaf (also in Eichards, Gramm., 1753, p. 60). Williams, in Dosp. Edeyrn, 823 and 1116: Dimet. y for a (probably meant to denote South Wales in general, as always in Davies), and in tbe dialectal lists: Gwentian mi y garaf .

[G9.] Sal., N. Test., uses y very often for a; c.g., pa beth y dderbyneist ac y glyweist; am y pethe y ddywedesit yddynt can y bugelydd, f. 83&; ir apostolon y ddetholesei ef, f. 170a; y gwyr y ddodesoch yn carchar, f. 177&; am y pethae hyn y ddywedwn, f. 178a; yno y gesodesont (yd anvonesont) wyr y ddywedent; ni y (== ei) clywsam ef yn llavaru . . . ., f. 179; etc. In Huets Gweledigacth leuan it occurs even more frequently. Cf. yr hon (sc. gweledigaeth) y rroedd dyw yddo ef yw ddangos yddy wasnaethwyr yrrein y orvydd yn vyan ddyfod y ben; ac y ddaugosoedd gan y angel yddy wasnaethwr loan (ac ef y ddanvonoedd); yr hwn y dystolaethoedd o eir dyw ac o dyst. I. Chr. . . . ac pob peth ar y weloedd ef, f. 373; ac ef y ddayth ac y gymerth y llyfr, f. 379; a phwy y ddychyn sefyll, f. 380; (= a phwy a ddichon sefyll, ed, 1873); happys ywr neb y ddarlleyo ar rrei y wrandawant geyryey y bryffydolaeth hon, f. 373, etc.

Also frequent in Addit. MS. 14,921, 16th cent. {John Maundevles Traves, Gwentian dialect); Addit. MS. 15,038, f. 7Sb: am y wneythoclii ero = Addit. MS. 14,973: am a wnaethochi erof; Lyfr Achau (Breconshire, 1002), Edward Kaer yn Arfon Yr Ayle y Bryudes


(delwedd B2732) (tudalen 151)


Elsbed, ii, p. 11: ar wraig hono y fyssai yn bricd a , p. 39; Tydwal . . . y fraethwyd yn y glyn . . ., p. 16, etc; Addit. MS. 14,973 (Rees Prichards, Poems, 1640): am bob gweithred ddrwg y wnelom ar gair ag y ddwettom, f. 102b; r maint y wueitho y, f. 65, etc; Caui. ofer y Cymry, 1672: mae yn vfFern fil o filiodd o wyr Ifaingc y bwrpassodd (a fwriadodd) yn eu henaint brudd difaru heb gael arfod wneuthur felly, p. 108; tl yn gywir am y gefaist, p. 173; a hyw n l y wers y ddyscont, p. 229; cynta peth y wuel dy ddwylo, p. 264, etc (see pp. 234, 261, 321, etc). In some of the manuscripts used in the critical edition of three of lorwerth Yynglwyds poems (in Y Cymmr., vol. vii), p. 185 (er a wnelom): B. 10, y nelon; (i alw a wnaid o lynn nedd): R, i naid; B 10, p. 187, Uyma gowydd y wnaeth I. V., p. 187 (pa vn a ofyn pa nwyf): B 10, y ofyn; p. 191 (a ddvg, relat.): R, i ddvg, L, y ddyg; (S 1 a helm a ddwg haul): L, helem ddwg hael; p. 193 (dw a ddowad): L, ddwad; L, er edn elwir dwyf, etc.

E. The Demonstratiye Pronouns.

[70.] Hyn, liynny, are said to be used in South Wales for masc. and fem. both. See Davies, Gramm. (Zeuss, ^p. 394), Y Traeth., iii, p. 10; WiUiams, Dosp. M., 1274. In Addit. MS. 14,921 (16th century, Gwent.), yr hwn (rwn) is used at the beginning of a relative sentence for sing. and plur. Cf. yr apokalyps yr hwn sydd yny bibyl, f. 3a; yr syttai .... yr hwn sydd . . . ., f. 3a; or hoU ilonds yr hwn oedd yddi hi, f. 4&; a ffob reliks rwn oedd, f. 21rt; y llethyr[e] yr hwn, f. 47rt; llawer o ilonde a thiroedd . . . . yr hw fydde ryhir draythy o hanynt, . 48. Also hwnnw: y tirredd hwnw, f. 13.

[71.] Hwna, hona, are of frequent occurrence in the modern language. They are formed after the model of yna, Hwno also occurs, but very seldom. Cf. Sal., JS. Test., hwnaw (pron. hwno, as he writes also ynow, o ddynow, f. 4a, 98a, 104, etc), f . 7, 40, etc.

[72.] These pronouns are variously altered, if they are used as proclitics or enclitics. Gweithian, noson, rŵan, are the results of gweith hon, nos hon, yr-awr hon. Cf. gweithon,


(delwedd B2733) (tudalen 152)


nunc, Zeuss, ^p. 618; gweithon and gweithian, Spurrells Didionary (now). Jwas most probably introduced from the plur. gweithiau, after the fact of gweithon, gweithan being a compound of gweith and hon had been forgotten. Noson, f. night time, certain night, one night, Sp., Dict

[73.] Yrŵan, rŵan, existed as early as in W. Salesburys time, who writes {N. Test.) yrwan, f. 352a. Thereore all other orthographies, apparently intermediate forms in Salesburys N. Test. and in some later manuscripts, are forms of no real existence, tending to reconcile in -wrriting the spoken rwan and the literary yr awr honn. It is, however, of interest to record them, as they clearly show how little in many instances apparently genuine and phonetic spellings in later manuscripts can be relied upon.

Cf . Sal., N. Test.: yn yr awr honrio, yr awr hon (ynawr; this is S.W.), f. 590; yr awrhon, f. ba; yr owrhon, f. 159a; ac owrhon hefyt, f. 325; yr owron (Huet), yr owhon, f. 235a, 239; yr owon (nawr), ff. 107!, 167, 175a; ynawr (yrowan), f. 103&; yrwau, f. 352a; Gr. Roberts, Athr. Gr.: yrowron, pref.; Y Drycli Christ.., yr owran, f. A2. Addit. MS. 14,986 (16th cent.): yn rawr hon, f. 20; yr awr hon, yr owron, ff. 20Z^, 35a; yrowan, ff. 1-h, 30, 32; yr wan, f. 29a. In this text, a religious drama, on which see my Beitr., p. 19, 5, the metre (verses of seven syllables) helps to establish the original (S. W.) text, cf. f. 27a: yr ywan wrth orchymyn vynhad i Iddwy vi yr ywan yn kodi, where nawr for the bisyllabic rŵan mnst be read. In other mauuscripts of this text: Addit. MS. 14,898, f. 70&, rowran; Addit. MS. 15,038, f. 62b, yr owau = rwau in 14,973 (yr owan, f. 5), etc. E. Lhuyd gives N. W. yrŵan, S. W. ynawr. All other sources tend to confirm this. In Sc-en Cpnrn, i, p. 192, a Northman is laughed at on account of liis rŵan. [74.] Hynny is an enclitic after prepositions, etc, and after y rhei, and is much shortened and altered in the col loquial language. Cf. Sercn Cymru: am ny, iii, p. 224; fel ny, iii, p. 186 (i6., fel ma, iii, p. 227; fel na, i, p. 162; y siop dma, Seren Gomer, xxxvi, p. 37); Yr. Arw., erbyn hni, 17, 7, 56, etc.


(delwedd B2734) (tudalen 153)


[75.] J. D. Tdỳs^Gramm., gives for the literaiy y rhei liynny (aud hyn): yr hain, rhain, yr heinei (?), yr heini, yr heiny, yr hai, rhai, rheieu (five times); or hai, or rhai, or rheieu. Davies, Gramm.: yr rhai ny, rhei ni; this was wrongly separated yr heini, hence heini barbare; cf. yngaser i (razor, Powel, Dimctian Loanwords, p. 29, etc.; ogla = arhoglau, rhoglau, Sweet, p. 431, etc). Gambold, Gramm., p. 94: rhei ni, rhei ny, rhai n. Sweet, Spokcn North IV., p. 442: rhgin, rheiny, rhaina.

Cf. Sal., N. T.: yr ei, f. 171 />; y rhe ini, f. llla; Huet: y rrcin, f. 385; yr rein, f. 382a. Athr. Grist.: yr heini, p. 33. Addit. IS. 14,921 (16th cent., Gwent. dialect): ri, f. bbh; yrhai, f. 13i; yrahai, ff. 106, 42; ar hai, f. 42/* (and), arhi, f. 596; or ri hyny, f. 29!; sef oedd raiui, ff. 40, bba; y rain. f. o^a; ac ar yrhaini, f . bba; ar raini, f. 39ẃ; ar haini, ff. 36a, 366, bba; arahini, f. blb; arhini, f . bbh; arhin oU, f . 9a (and). Addit. MS. 14,986 (16th cent.): yrheini, f. 126; barna hyrein i, f. 156.

[76.] In the colloquial lahguage y rhai hyny is also used after substantives, instead of hyny alone. Y dyddiau rhain, y munudau rhain, are quoted in Scrcn Gomcr, 1818, p. 328; y geiriau rhai n, Y Beirniad, iii (1862), p. 344. Cf. Rowland, Gramm., *p. 51; Cah. fcvj. Tomos, y briwars (Eng. brewers), rheini, etc.

[77.] Demonstrative pronouns are proclitics before some local adverbs: nacw = hwn acw (Sweet, p. 442), nwne = hwnw ene iu Merionethshire. Cf. Cah. fcio. Tomos, dwin meddwl y try nwne allan yn dderyn go dda, p. %Q (troi allan, to turn out); neno fy nain, Holyhead {Punch Cymracg, No. 3, p. 3) = hon eno. Cf. ene, Cah.fcw. T., pp. 18, 47; dene, pp. 7, 9, 345, etc. The counterparts of uacw, etc, are wnco, onco, ynco = hwn (y)co, hon (y)co, caused by the different position of the accent (hynco, D. S. Evans, Lhjthyr., index). Wnco, onco. are said to be S. W. (D. S. Evans),


(delwedd B2735) (tudalen 154)


r. The Interrogatiye and Indefinite Pronouns.

[78.] Pwy (cf. poebennac, MS. A. (Yenedot. Code), p. 41; puebennac, p. 42) is used in certain texts to denote things as well. In the BooJc of Herg. it relates to collective nouns, comprising a number of persons; e.g., col. 562, p6y y vydin burwenn racco; col. 763, y wybot p6y y niuer racco; but also col. 558, dywedy di ynni p6y dy lyssen6. Pwy is used almost exclusively (and pa given in the margin by the editor) in Gann. y Gijmrij, 1672; cf. pwy ddaioni, p. 254; pwy faint (pa), p. 256 j pwy Us (pa), pwy bethau (pa), p. 292; pwy wlad, p. 362; o bwy le ei ceisiwn (pa), p. 373; etc.

[79.] Pwy un, py un, become pwyn, pyn. Cf. pwyn yw hwn, a Sal., N. T., f. 95; pwyn a pha ryw wreic yw hon, f. 946; pwyn pynac a ddaw, f. 92&. Lewis Morris, Addit. MS. 14,944, f. 138, quotes from John ap Howel: Bun bengall bwynbynnag oedd; notice the cynghanedd. Addit. MS. 14,913, f. 58&: pyn a vo y klwỳ, a aỳ . In the Booh of Hcrg. occurs pun: pun wyt, col. 680 (pa un). *

[80.] Pa and py, in pa achaws, pa ryw, etc, are used apparently indiscriminately in a number of old texts. Both often occur close together. A special investigation of Middle Welsh manuscripts is needed with regard to this question. I wiU only note that in some parts of the Mab., larllcs y ffynn., Per. ah Efr., and K. ac Olwen, py is extraordinarily frequent, totally outweighing pa.

Cf. col. 637-652: py weidi y6 li6nn, py diaspedein y h6nn, py dr6c y6 hynny, py der6 itti, gofyn . . . py beth a holynt, etc, at least 10 times; col. 661-702: y wybot py gyfranc y6 yr eida6 ef, py gybellet odyma y6 y cruc, etc, 5 times; col. 811-814: py liuy di, py dr6c yssyd arnat ti, py ystyr, py hyt bynnac; and col. 831-837, 5 times, etc.

[81.] Paham, Zeuss, -p. 399 (pa-am), becomes pam; p(a)ham, p(h)am being the intermediate forms, as the fol-


(delwedd B2736) (tudalen 155)


lowing similar facts tend to sliow. Pabar (A., Yenedot. Code: pa har emae macht ae ar pelit mauur ay ar peth beccan, p. 56; pa/ar emae ebridiu ef ae ar pedeirarugein ae ar keuiauc, p. 64) = pa ar as pa am; pyr seems to be pa-yr, cf. gvae vi pir wuuf ar di kivuolv, B. of Carm., Skene, p. 4.3; guae ti din hewid, pir doduid imbid, il).; guae uinheu pir deuthoste imgotev, No. 6; pyr na threthch traethat, B. of Tal, No. 7.

[82.] Pa han: o ba han y daw ffydd, Sal. N. Test., f. 234 (whence). I suppose han to be identical with han- in han- fod, and both to be = ahan, ohon, the preposition oc with the obscure -*son- (see above, 41). Whether this -son- be of pronominal origin or not, at any rate ohon- is pre-Cymric, and the alteration of *cs into h and the use of ohon- with suffxed pers. pronouns necessarily destroyed the perception of its origin, and it was treated like an independent preposition. Theloss of the unaccented o is not irregular; cf. mae, nachaf, moes(?), gwneuthur, etc. This assumption will probably be strongly supported by a form in the oldest Yenedotian MS. A., if there be no error or misprint: pa hon emenno dystrihu etestion, p. 77 = bahan in MS. J., py ford in MS. D.

[83.] Pan (whence) is used in Middle Welsh texts in a peculiar manner at the commencement of an afrmative sentence, after it has been used regularly in a preceding direct or indirect interrogative sentence. Cf ba hid ei d a phan doit. Ban deuaw o caer Seon, B. of Carm., Skene, No. 35; B. of Hcrg., col. 886: a govyn ida6 o pa le pan hanoed; panhanyf i o freink (Bown o Hawtwn); col. 612: o ba le pann dathoed. ac o freinc panyf; col. 661: pa le pan deuy di; pan deuaf o Iys arthur; py le pan doei; pan deuaf o . . . . col. 665, etc.; Ll. Gw. Rhydd: . . . . o ba le pan dathoed; ac o ffranc pan wyf, p. 8; cf. pp. 124, 132, 142, 144, etc. This construction invoIves a kind of attraction, as it is called in classical grammar. It is impossible, on account of Ir. can (whence),


(delwedd B2737) (tudalen 156)


to explain pau in o ba le pan as pa han, like pam, etc, thongh all other circumstances are apparently in favour of this.

[84] Pawedd becomes in the colloquial language pwedd, bwedd; cf. bwedd y gallsei, Cann. y Cymry, 1672, p. 325. This illustrates the origin of pam from paham, and shows that pdd from pa fodd went through the intermediate form of *pfodd; cf. podd, Sal., N. Test. (I have lost the reference to the f.); pdd, J. D. Pthŷs, Gramm., p. 128; bodd y chwi = pa fodd a ydych chwi (Hughes, n Essay on . . . the Welsh Language, 1822, p. 30, from South Wales). Another example of an / apparently lost between vowels, but really dropped to facilitate the pronunciation, after a consonant (cf, tf, pf), is dd, from dyfod by the medium of *dfod. *Dfod was dialectally altered in another manner: dwad; the awkward collocation df being thus avoided also. Cf. also cyfodi, codi, and cwad, like dwad, from *cfod. Dwad occurs in Y Drych Christ., f. B 16, Cann. y C, p. 61, Williams Pant y Celyn, Yr Arweinydd, Cahanfei. Tomos, etc. On cwad, cf. Addit. MS. 15,038, f. 606: kwad i vyny = 14,973 (the same text), cwad i fynu; 15,059, f. 223. On w from/, cf. sgwarnog = ysgyfarnog, sgweny = ysgrifenu, cwarfod = cyfarfod, etc, in ISTorth Welsh dialects.

[85.] Pynnag in pwybynnag is a most interesfcing word, common to Welsh, Cornish, and Breton. There exists a Welsh dialectal form, cynnag, which is extremely seldom met with. I wiU first illustrate the forms and usage of pynnag by a few examples.

Cf. pa uacht panaac a vrthego, Yenedot. MS. A.^ see Zeuss, ^p. 400; L. {Dimet. Code), p. 198, pybaunac; S. (Addit. MS. 22,356), P6y bynaac syrhao y gilyd 6erin y pedeir glad hyun . . . taled pedeir bu, etc, f. 104. These are the only examples I know of jmuaac.

Tit. Jy. 2 = B. { Yenedot. Code): a phebeunac, f . ha {ib., enteu, heunỳ); Harl. 958= 2.; p6y bỳnhac hagen n boida6 etiued o gorff, f. 49!; kaus ỳmpeth buhac bo breint u yr vn rỳ6, f. 25a; B. of Herg.: a ph6y varcha6c bynhac a vynuei bot , py beth bynhac ar y fei o nerth yno (col. 175, 178); Addit. MS. 19,709:


(delwedd B2738) (tudalen 157)


py di bynac y bo haelder anianal ; Ll. Giv. Rhydd.: a pby gyhyt bynnac, p. 195; pa gyhyt bynnac, p. 173.

Sal., N. Test.: a phwy pynac ar nys derbyn chwi, f. 15a; and ( = ond) pynac a dyngo ir offrwrn , a pha tuy pynac ydd eloch iddo, f. 97a; am yr holl pethae bynac ar a roddeist imi, f. 161a; y sawl bynac na does ganthynt y ddusc hon, f. 376 (Huet). Addit. MS. 14,986: pwy bynac ag a ddysgo y kylandyr hwnn, f. Sb; Bardd., i, p. 334: bynnacryw o beth y bo hynny. [86.] The modern forms, which I am about to record, require an explanation in two particulars: bynna occurs besides bynnag, and bennag, benna, are found in South Welsh dialects. Cf. Cann. y Cymry, 1672: i beth bynnag fo duw n ddanfon, p. 206; and crd beth bynna ddweto r ffeiriad, p. 404; pwy bynna fytho, p. 238; ple bynnar elwyf, p. 324; pwy bynna geisio maddeu, p. 82, etc. Addit. MS. 14,973 (1640, S. W.): beth binnag; E. Lhuyd, Arch. Brit., pref.: pwy binnag; but Seren Cymru (Dimet. dialect): beth bena, iii, pp. 5, 45, etc.; pwy bena, iii, p. 186; beth benna (Aberdare), etc.; Medd. Myddf.: pwy achos benna ag y bo, 324; pwy benna gaffo, 187, etc,

As I think the -ac in pwybynnac (Old Welsh papedpinnac, patupinnac, Marc. Cap.) to be the preposition oc (ac, o, a), the occurrence of pwybynna is in no way surprising. Pwy- bynn- ac a is whosoever of (these) who; cf the common use of oc a, ag a, or a,ar a. In the Southern dialects the y in the penultima, etc, is pronounced rather as i (see my Bcitr., 64-74); and so pinnag is regularly written. But pennag, penna, is rather more common, and must be compared with pethewnos (and perhaps pesawl) in the same dialects. I cannot explain these anomalies.

[87.] In my Bcitr., pp. 78, 79, I gave all the instanccs that I know of cynnag. This is the only form (occurring nine times) used in the Gwentian MS. 14.921 {John Maundevilles Trarels). Cf pwy gynac elo, f. 8; ac phwyby gynng a ddarffe f. 11&; ac pwy wr kyn ac ele pa gyn pwy glefyd y fai arno . Tlh; pagyn fo n kydnbo i. 36a; a ffa rcli


(delwedd B2739) (tudalen 158)


pwy gyii fyno (= ofyno) yddi, f. 49a; a fetli bygync y dynese o hano, f. 50a; a cliyn pwy yfo, . 59a; pa gyna pa yfo, f. 596 (tlie fuller context is printed l. c.) The two bygynac are of course clerical errors, which show the tendency of the transcriber to use the dialectal form.^ I never met with cynnag in any other text; but Davies, Dictionary, has: pwybynnag, etc, et dicunt demetice Gynnag pwy, gynnag pa beth (he never separates the Dimetian from the Gwentian dialect). The same is to be found in Eichards Dict. (1753), and Spurrell too has: cynag, adv. soever.

[88.] The Cornish penag and the Bretonpennac show that pynnac (panaac, pynaac?) is a pre-Cymric form. If -ac is separated (see 86), the remaining *pan (pana-?), *pyn, is evidently identical with Sanskr. cana (in kacana) and hun in Got. hvashun (Skr. ni kacana = Got. ni hvashun). Cana contains ca, Gr. re (hence c from k), and the negative na. The Gothic -hun is the unaccented form, as un, the Gothic representative of unaccented *en (nasalis sonans), shows. In the modern German irgend the accented form is preserved. (Middle High Germ. iergen, from io wergin, o hwergin; hwargin, hwergin, anywhere, ags. hvergen = the local adverb hvar (Gothic), (Lithuan. kur) -h gn, g according to Verners law.) In the Brythonic languages pana- is exactly the Gothic hun, an being the equivalent of Gothic un, the nasalis sonans (cf. cant, dant, etc). Py- in *py na-, pyn-, Corn. and Bret. penn-, originated independently in each of the three languages by the shifting of the accent to the closely adjacent -c a, -c a (as li in the frequent Middle Welsh pynhac, as h in the conjunctives mynho and in the comparatives, etc, shows); cf the preposition am- ym-, etc

[89.] As to cynnag, I cannot explain it. Until more details are available as to its area of prevalence, and perhaps as to alterations of other pronouns in the respective local dialects,

1 I now hold by- to be rather py of py-gynnac, pa-gynnac [4, 7, 87].


(delwedd B2740) (tudalen 159)


it is equally easy and unconyincing to assume merely an assimilation of the consonants (cynnac like gangos for dangos?), or a corruption of pwy bynn aca and pwy aca into pwy g-ynnac a, or to connect c with cw, cwdd {B. of Tal., Skene, pp. 127, 145, 146, etc), on which see Ehŷs, Bev. Celt., vi, p. 57, seq. Cf. also Ascoli, Sprachwiss. Briefe, 1887, p. 165, n. 1.

[90.] On pwy, pw, py (from po-i) in pwy gilydd, etc, see Eliŷs, Bemic Celt., vi, p. 57, scq. A few examples are: or mor pygilyd, Mcib. (Guest), iii, p. 265; or mor bwygilid, Ll. Gvj. Bhydd., p. 21; kyut bwy kynt (sooner or later), p. 82; or mor pwỳ ỳ gilyd, Cleop. B. 5, f. 9Sa (cf. yw y?, 62); bygylydd, Sal, N. Test., f. 395; or iaith bigilidd, Gr. Eoberts, Gramm., p. 106; cf. L. Landav.: or carn dicilid, p. 183; Addit. MS. 19,709: or mor y gilyd, f. 14, etc

[91.] Peun-, and by a wrong separation peu-, was abstracted from the old peunydd, peunoeth (see Zeuss^, 618), and transferred to various other nouns, denoting every.

Cf. peiinos, Spurrells Dictionary; peutu (on each side, on both sides), ib. A dialectal difference is to be noted here. Y Traeth., iii, p. 13: N. W. oddeutu = S. W. o beutu. Cf. Medd. MydJfai: o beutu r wialen, p. 106; o beutu th liniau, 378; etc.; Seren Cymru . o beutu, i, p. 272; boutu, p.449; o boutu, ii, p. 226, iii, p. 306; o boitu, ii, p. 262; iii, p. 104; Gwent. dialect: boitu (Y Gwladgarwr, Aberdare, 2, 6, 1860), etc,

Beuneth, B. of Herg., col. 523. lolo Morganwg (Addit. MS. 15,003, f. 94) gives the foUowing South Welsh forms: beunos, beunydd, beuparth (everywhere, or in every part), beutu, beugilydd (everyone), beulin (every kind, species, genus), in which beu- replaces bob-. Cf. beuparth, Mcdd. Myddf, 439.

[92.] Let me, in conclusion, mention a North Welsh and a Dimetian idiom, neither of which has as yet been sufciently illustrated to permit an opinion on it to be formed.

Y Traeth., iii, p. 13: N. W. bod ag un, bod y pen = S. W.


(delwedd B2741) (tudalen 160)


bob un, bob yr un, bob pen, are given; cf. also D. S. Evans, Llythijr.: bod ag un (= bob un). These forms really occur in popular texts. Cf. Seren Cymru, i, p. 192: Y dyle ni bod y gun fyned y wrando fo, is said by a native of Merionethshire, whereupon a Dimetian, mocking him, says: Beth yw ystyr bod y gun, a rwan (see 73). The same Northman says (i, p. 292): Bron bod y gun or capeli acw rwan; and (i, p. 373): Yu agos bod y gun. Yr rweinydd (PwUheli j: ynd ni chyiff pob math o Twy hen flagiard (rhyw, blackguard, cf. bachgian) ddim loidgin (lodging) yno, ynd mi gan damad bod yg un hefud, 2, 10, 1856; Y Genedl Gymreig (Caernarvon): Pobol y set ( = society, cf. Powel, Dimet. Loan Words, N. W. seiat, pl. seide = S. W. seiet, pl. seiti) fawr ynte oeddan nhw bob yr uu (1885). Bod yr un seems to be a half dialectal, half literary (bob yr un) f orm; Cdban few. Tomos: mae on u nhabod bod ag un, p. 138.

[93.] The Dimetian peculiarity I wish to draw attention to is the use of w, which is said to occur at the end of short sentences, questions, or commands, but is not limited to them; e.g., bler ych chi yn mynd w? Dowch yma w ! I found it in Seren Gymru, ii, p. 105: fachgeu, ble rwyt ti yn mynd, w? It is unexplained.

May 2, 1887.


(delwedd B2742) (tudalen 161)



Professor Rhŷs, who has kindly aided in looking through the proofs of the above paper, has appended the following annotations:

P. 114, 11. 7, 8. They are only accented on the first syllable in Merioneth, as far as I know.

P. 114, 1. 16. Eh yr f, for which one may also hear hebrfi, has the yr wrongly separated from the rest of the word in mediaeval Welsh, as hebr(fi) is a survival of the deponent verb, corresponding to the Irish sechur, Latin sequor, aud used in the sense of I reply or I rejoin. That the supposed yr is not the definite article is seen from such instances as heb yr Arthur, says Arthur, which should have been written in mediaeval Welsh, hebyr Arthur, and in modern Welsh, hebr Arthur. Mediaeval Welsh hebyr had the sound of modern Welsh hebr.

P. 115, 1. 2. Mi: this is used in all the dialects, I believe. It is at any rate as common in South Wales as in the North. It is also pronounced me, and it is nothing but fe or fi put into a quasi-radical form,

P. 116, 11. 7, 8, 9. It is also used in South Wales.

P. 153,1. 6 from the bottom. Neno fy nain stands for yn enw fy nan, in my grandmothers name. The demonstrative in Gwynedd is nwna, fem. nona.

P. 153, last 1. In N. Cardiganshire the forms always are hwnco, fem. honco, with an h.

P. 160, 1. 2. Bod agun seems to stand for bb af n, or bawb ag n, each and every, or more literally, every and each.

P. 160, 11. 12, 14. Y st fawr is well known to be the big seat in which the elders in a Nonconformist chapel are wont to sit around the pulpit; it has nothing to do with seiet or seiat.






a A / / e E / ɛ Ɛ / i I / o O / u U / w W / y Y /
ā Ā / ǣ Ǣ / ē Ē /
ɛ̄ Ɛ̄ / ī Ī / ō Ō / ū Ū / w̄ W̄ / ȳ Ȳ /
ă Ă / ĕ Ĕ / ĭ Ĭ / ŏ Ŏ / ŭ Ŭ /
ˡ ɑ ɑˑ aˑ a: / : / e eˑe: / ɛ ɛ: / ɪ iˑ i: / ɔ oˑ o: / ʊ uˑ u: / ə / ʌ /
ẅ Ẅ / ẃ Ẃ / ẁ Ẁ / ŵ Ŵ /
ŷ Ŷ / ỳ Ỳ / / ɥ
ˡ ɬ ŋ ʃ ʧ θ ʒ ʤ / aɪ ɔɪ əɪ uɪ ɪʊ aʊ ɛʊ əʊ /

ә ʌ ẃ ă ĕ ĭ ŏ ŭ ẅ ẃ ẁ Ẁ ŵ ŷ ỳ Ỳ

gw_gytseiniol_050908yn 0399j_i_gytseiniol_050908aaith δ δ
wikipedia, scriptsource. org




Ffynhonnell / Font / Source:

Creuwyd / Creada/ Created: 22-11-2017

Adolygiadau diweddaraf / Darreres actualitzacions / Latest updates:

Delweddau / Imatges / Images:



Archwiliwch y wefan hon
Adeiladwaith y wefan
Beth sydd yn newydd?

Bler wyf i? Yr ych chin ymwld ag un o dudalennaur Wefan CYMRU-CATALONIA
On sc? Esteu visitant una pgina de la Web CYMRU-CATALONIA (= Galles-Catalunya)
Where am I? You are visiting a page from the CYMRU-CATALONIA (= Wales-Catalonia) Website
Weə-r m ai? Yu a-r vziting ə peij frm dhə CYMRU-CATALONIA (= Weilz-Katəluniə) Wbsait

Free counter and web stats