(delwedd 7898b)



(delwedd 7375)



Wedi du cywiro
Heb eu cywiro

56:: 57::
58:: 59:: 60:: 61:: 62:: 63:: 64:: 65:: 66:: 67:: 68:: 69:: 70::

71:: 72:: 73:: 74:: 75:: 76:: 77:: 78:: 79:: 80:: 81:: 82:: 83:: 84:: 85::

86:: 87:: 88:: 89:: 90:: 91:: 92:: 93:: 94:: 95:: 96:: 97:: 98:: 99:: 100::

101:: 102:: 103:: 104:: 105:: 106:: 107:: 108:: 109:: 110:: 111:: 112::

113:: 114:: 115:: 116:: 117:: 118:: 119::

259:: 260:: 261:: 262:: 263:: 264:: 265:: 266:: 267:: 268:: 269::

270:: 271:: 272:: 273:: 274:: 275:: 276:: 277:: 278:: 279:: 280::

281:: 282:: 283:: 284:: 285:: 286:: 287:: 288:: 289:: 290:: 291:: 292::

293:: 294:: 295:: 296:: 297:: 298:: 299:: 300:: 301:: 302:: 303:: 304

VOL. IX. 1888.



The present paper has been compiled upon the
methods used in my articles on the Welsh vowels
(Beitrge zur cymrischen Grammatik, 1), in that
upon the Welsh consonants (Revue Celtique, to
appear in Jan. 1888), and in that upon the Welsh
pronouns (Y Cymmr., vol. viii). The study of the
Welsh verb is attended with greater difficulties,
both internal and external, than that of the
pronouns; for while these isolated words can
easily be traced through the successive periods of
the language, we have, in the case of every verb,
to distinguish between the characteristic form of
its stem and the verbal terminations. The
syllables forming the latter were dropped
in Welsh, according to the laws governing the
phonetic treatment of final syllables, and the
actually existing endings are the result of the
different sounds characteristic of the stems,
modified by the influence of the lost original
terminations. A classification of verbs is
therefore the first requisite, and the solution of
this problem will, in all likelihood, be furnished
by the variety of terminations still existing in
some parts of the verb. Cf. 3rd sing. Pres. -awt,
-it; Pr. Sec. -ei, -i, -awd; s-Aor. -es, -as, -is;
plur., -assom, -yssom, -som; pass. -ir,
-awr; Part Pret. Pass. -et, -it, -at; etc, But by
the analogical prevalence of some of these endings
the scheme of the Welsh verb has been, during the
history of the language, reduced to such
a degree of uniformity that only by a full
collection of Middle-Welsh materials, which are at
present accessible only to a


limited extent, coupled with an equally methodical
scrutiny of the Cornish and Breton languages,
could the necessary foundation for further
procedure be obtained. For this task the materials
at present at my disposition are inadequate. I
propose, however, in the following paragraphs to
review some other less noticed questions
respecting the Welsh verb.

Evander Evans's Studies in Welsh Philology, and
Rhŷs's article on the Verb in the Revue Celtique,
vol. vi, are the most useful papers extant dealing
with this portion of Welsh grammar.

The arrangement of these paragraphs is made to
conform to that in the Grammatica Celtica (pp. 2/


1. Evander Evans (Studies in Cymr. Philology, 13)
first recognised -ydd as a termination of the 2nd
pers. sing., corresponding to the regular Cornish
(-yth, eth) and Breton (-ez) terminations:

pyr nam dywedyd, B. of Tal., p. 145;

ti a nodyd a rygeryd o pop karchar, p. 180;
truan a chwedyl a dywedyd, B. of Herg., p. 231
(Skene); etc.

Also in old proverbs: gwell nag nac addaw ni

I found in Add. MS. 14,921 (16th cent.), modise
fidei quāre dubitasti .... hyny yw tydi o ffydd
wan pam y ofnydd (corrected by another hand into
ofnydi). (1)

Davies (Gramm., 1621) gives cery as a Dimetian and
poetical form for ceri; and Rhŷs (Rev. Celt,, vi)
restores diwedy in B. of Carm., p. 57, by means of
the rhyme.

I suppose -i and -ydd to be the endings proper to
the verbal j-stems; they are doublets, like -i and
-ydd, -edd, in the nominal jo- and jā- stems,
caused by different accentuation, and contain the
secondary ending -es, like Ir. asbr. *Ber-es

(1) E. Evans (Stud., 16) gives from Huw Llwyd of
Cynfal (Cymru Fu, p. 352, a book which I do not

know): nac a ofuith moi gefnu (whose desertion
thou wilt not fear), which he holds to be the 2nd
sing. Opt. in -yth instead of -ych.


would have given *ber, like *beret, and was
probably lost by this coincidence. In the later
language cari is used, like carwch, tarawch, for
cerwch, tarewch, etc. (Rowlands, Gramm., 2/216).

2. The various existing types of the 3rd sing.,
using the same verb as an example, are: dodid,
*dodod, dod, dyd, *dodo, doda, dodiff, dodith. -id
and -od (-it, -awt, -ot) are those which present
the greatest interest. They are known from

the old Brythonic glosses: cf. istlinnit, Juv.;
dodiprit, Lux.; doguolouit, Orl.; crihot, Lux.;
cospitiot, fleriot, Orl. Evander

Evans (Studies, 5, 14, 19) gives numerous
examples of -it and -awt from Middle-Welsh texts;
cf. also B. of Carm., 4: chwerdyt bryt 6rth a garo;
dyg6ydyt gla6 o awyr; megyt tristit lleturyt llwyr;
chwennelcyt meuyl ma6reir; 29: ottid, tohid,
gulichid, gwasgarawt, llwyprawd, etc. (Evans).

Evans's most interesting remark is: "The Irish -id
of the 3rd sing. Pres. Ind. Act. is not used in
subjoined verbs, that is, in verbs following
certain particles, among which are the negatives
ni and na and ro; this idiom obtains also in
Welsh" (14, 1873). He goes on to quote sentences
like the proverb: trengid golud, ni threing molud
(Myv. Arch., iii, p. 177), and says that he found
-id only in absolute verbs. The Ir. 2nd and 3rd
sing., asbir beri, asbeir berid contain
different sets of endings: *beres *beresi, beret
*bereti (see Beitr. zur vergl. Sprachf, viii, p.
450). So we have in Welsh: cymmerid (*bereti),
carod (*carāti), and cymnier (*beret).

The subjoined forms of the simple verb are used in
Irisb, as in Welsh, if the verb is enclitic, after
the negations, etc. This accounts for the
disappearance of the absolute forms, which
were analogically supplanted by the subjoined ones.

3. Very curious are the examples given by
Evans(19) for the3rd sing. of the s-Pret, (-
essit, -yssit, -sit), which, in

the examples given by him, have to all appearance
the meaning of the active (pregetbyssit, kewssit,



llygrassyd, etc). These forms are evidently based upon the secondary preterite (dalyassei, cawssei); but they may have sprung from an analogical imitation of dodid: ni dyd, by *cewssid: ni cafas (?). The present in -id is often translated as an imperative. Davies (Gramm.) has: imp. cared, non nusquam cerid: cerid duw fi. The proverb: chwareid mb noeth, ni chwery mb newynawg (Myv. Arch., (2) p. 843), is given in Y Traeth,, iii, as hwareuid m. n. in South Wales, and chwareued m. n. in North Wales. In proverbs -id often survives; cf. Hengwrt MS. 202, Prov. 64 (14th cent.): dighit rỳwan elit rgadarn == Myv. Arch., 2 p. 843: diengid gwan, elid rhygardarn; ib.: elit ysguba6r gan dryc torth = Myv. Arch., (2) p. 845: elid ysgubor gan ddrygdoith, etc. On -aw, -o,
see Evans, Stud., 17.

4. The following verbs ordinarily form the 3rd

sing. (subjoined form) by so-called inflection of

their vowel or vowels. The following enumeration

is probably very incomplete.

A: pair, gaill, pairch, llaim, saiff, ceidw, geilw,

lleinw, teifl, deil, ymeifl; efe a deaill (Barddas,

i, p. 30). E: gwyl. O: try, ffy, cly, cny, tyrr,

rhy rhydd, gylch, llysg, trych, cyll, dyd

(gwrthyd), cysyd, deffry, cyffry, dicbyn (Davies);

a ddiylch, Add. MS. 14,973, f. 106b (dieylch, Casgl. Didr., f. 520a). The inflection extends over more than one syllable (1): envyn, B. of Herg., col. 1101; erys, ervjll,
Williams, Hgt. MS., ii, p. 332; gedy, tery, eddy, gwerendy (gadaw, taraw, addaw, gwarandaw); ef a edy, Y S. Gr., 192; and edeu, B. of Herg., cols. 632, 823; yd emedeu, Ll. Gw. Rh., p. 151; a
dereu, pp. 53, 232; gwerendeu, p. 36; a warendeu, p. 85, etc.; ettyl, gweryd (gwared), derllyn, merchyg, gweheirdd, ni wesnyth, ettib (atteb), Add. MS. 15,059, f. 209b.


(1) Cf. from Ll. Gw. Rhydd.: y gwetwery, p. 72; a phan vyrryyssynt hwy, p. 203 (bwrw); gwerendewch, p. 218; pann ymwehenynt, p. 52; gwerendewis, p. 110; yr honn y gwyssyneitheisti yn llauuryus, p. 149. MS. Cleop., B 5: gwesseneithyt, f. 114b; pei gedessyt yn vew, f. 129a; a werchetwys, f. 135b. MS. Tit, D 2: e gueheneyst ty, f. 48b. Cf. aso
oldBreton ercentbidite gl. notabis (Bern,), from arganfod.


Geill represents *galjet; these j- stems are not

to be confounded with the derivative j- stems, of

which the 2nd sing. gwyddydd is formed. A part of

them are old; cf. Breton quell W. cyll, etc.

The formation of this person and that of the 3rd

sing. s-Pret., the Part. Pret. Pass., and in a

certain degree the 2nd preterite, form sets common

to many verbs, and will permit the establishment

of verbal classes, if the materials

are fully collected. For there are numerous

neologies and dialectal preferences for certain

endings, and the same will more or less be the

case in Cornish and Breton.

So we have dyd dodes doded;

llysg llosges- llosged; etc,

but geilw gelwis gelwid;

ceidw cedwis cedwid;

teifl tewlis tewlid; caiff

cafas cafad; etc.

5. These old inflected formations (dyd) are dying

out, and in the modern language the types cara and

ceriff, cerith, are common. -a is taken from the

derivative verbs in f (-aaf *-agaf, Ir. -agim);

see Evans, Stud., 15. -iff is printed in

Aneurin and Taliessin (gogwneif hesslhut gwgnei

[leg. gogwnei?] gereint, "posterity will

accomplish what Gereint would have done", B. of

An., p. 89; ef g6neif beird byt yn llawen, B. of

Tal, No. 37 = Myv. Arch., 2p. 52a); but I

have not found it in Middle-Welsh texts.

Salesbury, New. Test.: amylhaiff, f. 39a; aiff, f. 84a; gwnaiff, f. 1576.

Gwel. Ieuan: nyd eiff ef, f. 377a; efo eiff, f. 387b; ef eiff, f. 392b, etc.

Griffith Roberts, Gramm., p. 60 (262): ceriph ne car.

Y Drych Christ.: efa wanheiph ag eph yn lesc, Bl; ni edewiph, B2b; ef a wneiph, Clb, etc.

Some writers use it excessively often, like Charles Edwards in Hanes y Ffydd
(1677); but in the literary language it is avoided.

In modern dialects -iff (and -ith) are common.

Cf. Davies, Gramm,: jam dudum vulgo ceriff, periff, rhoddiff (following ceiff).

Richards, Gramm.: sefiff, torriff, lleddiff.

Iolo Mss.: fe a wnaeff, p. 283; a wnaiff, p. 284; a ddielyff, ib.

Add. MS. 14,979 (17th cent.), f173a. ni wnaeff ddim yn i amser, . . . ni
wnaiff ddim . . .;

E. Llwyd: pwy binnag a edrichif (Arch. Brit., Pref.). Add. MS. 15,005, f32b:
lleddiff; f. 49a: hi ddwidiff, etc.

Hope, Cyf. ir Cymro, 1765: os misiff ddwad, p. viii


("if he fails to come").

Lex Cornnbrit., p. 12lb (North-Welsh): dyff

or deyff for daw.

Yr Arw. (Pwllheli): mi yiff, 17, 7, 56; y

gnyiff o (ni chyiff, 2, 10, 56).

Y Bed. (Monmouthsh.): fe leiciff, viii, p. 107, etc. (1)

6. I have found -ith only in texts of this century. Rowlands, Gramm., 4/80,
has the colloquial endings -iff, -ith, or -yth.

Cf. Yr Arw: dyudith (says), sonith, gofynith, 20, 1, 59.

Cab few. Tom.: mi eith, p. 33 (=fe ); gneith, p. 46; os na newidith o i
ffordd, p. 33; ne mi 'drychith o ar f 'ol i, p. 106; thewith hi byth yn

dragowydd (ni th.), a mi 'ddawith ddwad rwsnos nesa, p. 137.

Ser. Cymru: na cheith, i, p. 252; os na ofalith e, i, p. 272; cheith (ni ch.),
ib.; etc.

7. -iff has heen explained as an erroneous ahstraction of a termination from
ceiff. Caffael, cafael, cael, exist; from the two sets, caffaf, ceffi, ceiff,
and cf, cei, c, cei and ceiff
were selected; and in consequence of the preference for these two forms, eiff,
gwneiff, parheiff were first analogically formed,

and afterwards -iff was transported even to car:

ceriff. This is clearly proved by the presence of

e for a (ceriff: ceri, car),
which could not have heen caused by infection at

so late a date, but is the
old infected e of the 2nd sing. Seiff, from sefyll,

sefyd, is caused by the
coexistence of caffaf and cafaf. -ith may be a

phonetic change of -iff; cf.
benffyg benthyg; dattod dathod daffod;

dethol deffol. But if it be
genuine, as it can of course by no means represent

an old dental ending, it may have been taken from aeth, gwnaeth,

(1) In the MS. A of the Laws (ed. Owen) gwataf is often printed for the 3rd sing.
Cf. pp. 501, 506, 507, 509, 510, 527: os gwaaf tyst . . . .;

onys gwataf yr amddifynwr yr krair . . . .; ony wataf yr a. yr kr. . . .; ac o
gwataf ef hyny . . . .; o gwataf ef, etc. In this MS. f and ff are used for
both f (v) and ff (f): a fa le ymaen, p. 502; eff, wyff, p. 497; addeff, p.
509. There occur also wyntef, p. 528; atof, p. 523 (= atto); weydy hyny, p.
529; argloyd, p. 527; yewn, p. 524. Gwataf, if not, in spite of its frequent
recurrence, an error, is either gwata (f not pro-

nonnced) or gwataff (cf. eff), a form not elsewhere met with. Perhaps -af is
written for eff, and this for -eiff, a in final syllables being pronounced e
in certain dialects; cf. oedren, anhowddger, rwen, rheitiech, etc, in the late
Powys. Add. MS. 15,005.


etc.; not directly, since they are preterite, but, as I suppose, the
alteration of -iff into -ith was caused by the influence of aeth, etc,
pronounced aith or th, iff and the erroneously detached *-ith of aith bein^
both rearded as real terinina- tions, and -ith thns partly replacing -iff.
Even ceiff is ceith in dialects; see 6.


8. There is a growing tendency in Middle-Welsh to

supplant the terminations of the optative by those of the conjunctive, i.e, oe
(oy) and wy in the lst and 3rd sing. and 3rd plur. by o, with the exception
only of the lst sing., in which

-wyf and -of are both used. Oe occurs only in the oldest MSS., and in the
North- Welsh Laws.

Cf. (lst sing.) MS. A, p. 58: nydoes kenyvi atalloef ycgnic (sic, to you)
namin vimare ahunnu nys taluaf (sic) ycgui ac nis gustlaf. (3rd sing.)

B. of Carm., 18: creddoe (guledchuỳ, 16;

dirchafuỳ, 18, etc.

B. of Tal., 18: molh6y, roth6y, 19). 3rd plur.

-oent occurs very often in MS. A: eny kafoent, p. 5 (gymeront, MS. D); pan
uenoent, p. 10 (pan y mynho, D); palebennac edemkafoent er efeyryat ar dystein
ar enat (ymgaffo, D), pan ranoent er

anreyth, p. 16; ar e gladoet ackauarfoent ac Aruon, p. 50; bed a deuetoent, p.
389 (dywett6ynt, G, U); dim or a deuetoent (dywet6ynt, G, U); guedy
edemdauoent, p. 397; kinguibot bet a dewedoent, p. 73 (beth a dwettoent B. D.
.; dwettwnt C), etc.

(pp. 51, 55, 65, 74, 127, etc).

In MS. Tit., D2 (= B) -oent seems always to be used. Cf. ac na wnelhoent dm
namn can e gghor, f. 4a; pan gmerhoent, pan venhoent, f. 5a; pan ranhoent
hv, f. 7b; pa le bennac ed emgaffoent, f. 15a; mal e delehoent, f. 21b; pan
emchuelhoent, f. 60b, etc. (more than 24 times).

MS. Calig., A3 (=C): Ac o gwed e datkanoỳnt daw ef; hỳt en e lle e deloent
heprwng, f. 168b; ket anawoent, f. 198a; pan delhoent, f. 190b; na
phlyccoent, f. 191b; a kmeroent, f. 194b;

hỳt e delwỳnt pellaf, f. 169a; mal e delwnt, f. I73a; kaffwỳnt, f. 178a,

Addit. MS. 14,931 (= E): pan unhont, f. 4a; pan rannoent, f. 6a; pale bỳnnac
dmgafont, f. 12a (= B, f. 15a); a gfarfoent, f. 17a; val ỳ mnnont
talhoent cmeront, f. 21a; lle dlont, f. 21b; ket as dcoent, f. 24a;
pan deloỳnt, f.30a; pan elhont. etc - na wnelh6nt, f. 3a; ht dlnnt, d
ergythunt, etc.


MS. II (Owen, Laws): dylyoent, dyloent, galloent, dy wettoent,
pp. 736, 741, 760, 768.

9. -wynt is comnion in the South-Welsh recensions of
the Laws, and in Southern Middle-Welsh MSS. generally.

Cf . MS. L (Dimet. Code): ac yn g6randa6 ygneit a delh6ynt or
6lat yr llys, p. 180 (delhont, six other MSS.); yr hynn auo petrus
gantunt ac a vynh6ynt . . . . y amlyccau, p. 180 (uỳnhont, MSS.
N, P, Q), p. 199, etc. MS. Harl. 958 (= T): a wnelhỳnt, f. Yla.
MS. Cleop., A14 (=W): pan fohỳnt or wlat, f. 496, etc. Ll. Gw.
Rhydd.: val y caff wynt wynteu, p. 98; a arwydocaant y lcenedloed
a delwyntrac llaw, p. 277; pona coffawynt, p. 279; llawer o betheu
ereill a deloynt, p. 283 (doyn, p. 276; moy, p. 279). MS. C'leop.,
B5: a vrefwnt, f. 63; a vỳnhwnt hw, etc. MS. Jes. Coll.
141: niegys na allwynt adnabot. . . , f. 144, etc. (aNorthern MS.)

10. -of (not in Zeuss, p. 2 512): Ll. Gw. Rhydd.: yny
gysgof, p. 137; or a ouynnof udunt, p. 260. MS. Tit, D22:
a wnelof i, f. llh. In the modern lauguage o is introduced
in all persons (-of, -ot, -o, -om, och, -ont). -wyf is still used
(e.g., gallwi, Cann. y Cymry, 1672, p. 481; byth na delw i,
Seren Cymru, ii, p. 505; tra bydw 'i 'n gneyd, Cab. fcw. T.,
etc). In Ghrtmmars (Rowlands, Williams ab Ithel) carwyf,
-wyt, -wy, -y w, -ym, -wym, -ych, -wych, -ynt, -wynt are given
(used as Pres. Indic.). I doubt whether these forms have
any real existence; ic will be remembered that tlie same
forms of wyf (I am) are sufnxed to all pronouns instead of
the older endings (-of, -af, -yf).

11. -o- is certainly the reflex of the * of the conjunc-
tive, and oe has always been referred to the optative; but the
phonetic proof of this is very difficult. Oe (later wy) may
be *(i), the Cornish and Breton -i- *-- of the non-thematic
optative, and both may have exceeded their proper places
(the sing. and the plur.) (?). -i- is not wanting in Welsh.
Zeuss, p. 2 583, gives gwell gwneif a thi (melius faciam erga
te), Aneurin, p. 62 (ed. Williams), without adding any note;
ni bydif ym dirwen, B. of Tal, 31, 32, 33, 34 (ny bydaf,39);
acos ytydif ym gwen, 37, Myv. Arch., 2 $. 506;;t minheu


bydif, B. of TaL, Ruhn's Zeitschrift, 28, p. 91, n. 3. To these
forms belongs also the 2nd sing. in -yt (Zeuss, 2 p. 512) and
-ych, with cli of the pers. pron., 2nd pers. plur. We may
assume tliat at the time when the old -t of the 2nd plur. was
replaced by ch, the same was analogically done with
the more recent -t of the 2nd sing. (from ti). In certain
cases, indeed, if two forms are identical, though of different
origin, the rational change of one of them may cause a
merely analogical change of the other one; and besides, in
this case both are forms of the 2nd person. Davies, Gramm.,
gives cerych and carech; Eowlands and Spurrell, dysgych,
-ech, -ot ; Ll. Gw. Rh. } a wyppech di, p. 214; Addit. MS.
31,056 (17th cent.), pen fwriech di nhw ymaith, f. Wb. -ech
is caused by the coexistence of carech and cerych in the 2nd
plur. of the secondary preterite, an example of the kind of
analogy mentioned above.

0. Secondary Present.

12. 2nd sing.: Davies, Gramm., says: carit amabas, cerit

amares, poet. -ut, -yd; semper fere -ud. -et is the ordinary

form of the modern dialects, influenced by the 2nd plur. -ech.

In tlie plural both -ym, -ych, -ynt and -em, -ech, -ent exist.

Cf. Cann, y C, 1672; pe caet ti, p. 406. Addit. MS. 14,973:

letty a gavd pe raedred i ofyn, f. 105/y. Yr Arw,: mi fasat, ouddat,

mi rouddat (yroeddet), 26, 2, 57; pen ouddat ti, 11, 12, 56, etc.

(e becoming a in final syllables in the Venedotian dialects). Cal>.

few. Tom.: roeddet ti, p. 22; na baset ti; osgallset ti fgallasswn);

mi gowset ti, p. 30 (cawsswn); ni chlywset ti, p. 61. Scrcn Cymru:

cymeraset ti, ii, p. 47; pam na sharadet ti, p. 146; ceset ti, p. 243


13. The 3rd sing. offers similar problems to the 3rd sing.

I'res., for there exist -ei, -i, on one side, and -ad on the other,

wliich cannot have had the same termination. -i was first

recognised by Evander Evans (Studies in Cymr. Piil, 26),

and has been further discussed by Rhŷs (Ecv. Celt., vi). Doi

and cai occur in Middle-Welsh prose texts; cf. B. of Herg.: pa


doi arnat ti, col. 759; Ll. Gw. Rhydd.: na doi ef, p. 130; y doi
ynteu, p. 154; a doi, p. 154; ny doy, p. 207; y tygawd ynteu
y niynnei ef hihi oe hanuod. canys cai oc eu bod, p. 154.
-ei becomes at an early date -e in the colloquial language, as
is proved by orthographies like na dwettev neb onadunt
(Cleop., B 5, f. 6b), hadarnhaeu (S., p. 599, Laios), a gaffeu
(p. 594), peu veu (p. 597), -eu and -ei being at this period
(15th cent.) both pronounced -e, and therefore liable to be
exchanged. J. D. Ehys, Gramm., p. 128, has gall. Davies
says: vulgo profertur care, carase, but not before this century,
an obseiwation renderecl doubtful to me by the orthographies
above given.

14. -ad, -iad, is restricted to a few verbs. Bwyad, oedd-
yad, gwyddyad, adwaenyad, pieuad occur; see Zeuss, 2 p. 002,
Ehŷs, Rev. Celt., vi, p. 47?t. Cf. B. of Eerg.: ymbyat, cols.
1221, 1223 (Talhaearn Brydyd Mawr, flor. 1380); ny wydyat
hi, col. 1108, etc.; Ll. Giu. Rh.: a racatwaenat, p. 5; nysgwy-
dyat, p. 20; Cleop., B 5: ỳ gwidiat, f. 24; na wydiat, ff. 32fr,
376; na wỳdat, f. 37. Rhŷs gives gwyddad as a South-
Welsh form (N. W. gwyddai). It is almost fche only one of
these forms found in later texts; cf. Add. MS. 12,193 (1510):
nevpwy a wyddiad achos, f. 13&; na wyddiat, f. 22 (Trans-
lation of Rolewinch); Salesbury, N. T.: canys ef a wyddiat
f. 46; pe gwyddiat, f. 40, etc.; canys adwaeniacl ef hwy oll,
f. 134; also in Add. MS. 14,921 (Translation of Maunde-
mlle); Add. MS. 31,055 (ni wyddiad ddrwc dros dda, f. 34);
Cann. y Cymry; etc. Y (j) appears also before other verbal
endiugs, especially often in Y Seint Greal.

Cf. gwydyat (more than 30 times); a wydywn, pp. 231, 243, 300,
367; oedywn, 57; yr oedywn, pp. 239, 340, 422; ny dathoedywn i,
48; val yd oedewch chwi, p. 386; ti a aethyost, 13, 19, 44;
gwnaethyost, 12, 18; a aethyant, 7, 15; a doethyant, 10, 15;
a wnaethyant, 2; pei at vydewch, p. 143; a phei bydewch, 56.
Y is sometimes superfluously written in this MS.; cf. aelyodeu
(limbs), 2; twrneimyeint, 20, 21; mi a wasanaethyeis, haedyeist,
15, 19, etc.



In other MSS.: BooJc of Tal., 5: ny wydyem and wydem;
B. of An.: gwdei, 69, 19; B. of Ecrg.: ydaethyant, col. 1094;
L. G-i. Rh.: doethywch, p. 260 (kawssodyat lcg. kawssoydat?
cawssoedat, p. 224); Hgt., 59; gydynt, R. C, viii, p. 9.

15. The explanation of these endings is still extremely
uncertain. -iad is considered as belonging to the -stems,
and its d(t) and the absence of any infection justify the
assumption of the ending -to. -i and -ei may have only lost
a syllable ending in t, otherwise t would have been kept; so
they probably ended in *-e-t, and -i, -ei are the remains of the
characteristics of the stem with the derivative j'. -i is to be
explained as the -i of the 2nd sing. (see 1); -ei is regarded
by Bhŷs as the ending of the -stems; I should take it to be
the reflex of *-j-et. The existence of the doublets -i and
-ydd in the 2nd sing. Pres. enables us to concede the possi-
bility of such forms having existed in this termination, and I
rcally think -awd, -odd, to be a relic of them. Stohes first
explained -odd from *-jet; -ei is either *-jet, which I do not
believe, or *-jet.

16. -odd is replaced by -oedd from, as far as I can see,
the 15th century in South "Welsh, or more accurate.ly, perhaps,
in Gwentian texts only. Cf. from the Boolc of Trcv Alun,
near St. Asaph, written by Gutyn Owain: a gynnalioed, p.
629 (twice); Owen, Laws: a camgynnaliocdd, a lithroedd,
pp. 629, 630. In Salesbury's N. T., -awdd, -odd, and -oedd are
indiscriminately used, but in the part translated by Huet
-oedd, -oydd predominate. In the Gwentian MS. Add. 14,921
-oedd is the common ending, several times curiously written
-edd; Llyfr Achau (lreconshire), 1604: priodoedd, p. 8; a
wleduchoedd, p. 63, etc.; Eomil., 1606: pan weddioedd hi, ii,
p. 267. Owen Pughe onlinarily writes -ocdd, as does the
periodical, Y Grcal, 1806-9. In vol. i of Scrcn Gomcr (1814,
fol.) several columns are filled with letters and controversies
on -oedd and -odd, but I have not found any dialectal remarhs


of any service in them. -oedd is of course caused by oedd

(was) in such combinations as aethoedd, dathoedd, gwnaeth-

oedd, with the meaning of active preterites.

17. The plur. endings -ym, -ych, -ynt (-int), and -em, -ech,
-ent, are so far obscure as to fail to show by their phonology
to what classes of stems their difference is due. They seeni
to correspond to the Irish second and third series. From
the 15th century -eint, -aint, occur frequently in the 3rd plur.,
especially in later Gwentian texts.

Cf. Ll. Giu. Rh.\ traweint, p. 184 (trewis, p. 185; tereu, 3rd sing.);
a lauuryeint, p. 213. MS. Cleop., B5: aethasseint (sec. -Pret.);
MS. Tit., D22: tr6 yrei ybuasseint, f. la; ary messur y buasseint,
f. 8a; ysawl a yniardelweint; etc. Medd. Myddfai: a ddylaint and
a ddylynt, p. 296; a ddywedaint (Owen, Laics, p. 661, written
1685 in Glamorgan). Barddas, i: ag a ddodaint, p. 40;
oeddeint, p. 32; gweddaint, p. 42; a elwaint, p. 52; gwerthaint,
p. 64; medraint, gadwaint, etc. Davies says: poetice -aint and

These endings seem to be borrowed from the verbs in -f
(*-agaf), like the -a of the 3rd sing. Pres. Cf. a uuchocceynt,
Ll. Giv. Rhyclcl., p. 191; a wneynt, B. cf Herg., col. 720, etc.

D. The /S'-Peeterite.

18. The Ist and 2nd sing. -eis, -eist, are altered in the
modern dialects according to the phonetic changes proper to
the vowels and diphthongs of final syllables. So we find -es,
-est, in South Wales, except in the Eastern Gwentian dialects.
Cf. Salesbury, JV. T.: mi a weles, mi wyles, f. 378b; ti y
creest, ib. Scr. Gymru: gadewes I, iii, p. 226; mi gwmres,
p. 524; pwy welest ti, ii, p. 364; pan glwest, p. 524; wedest
ti, i, p. 232 (dywedaist), etc. Y Tyw. ar Gymr.: ond wetas
I, i, p. 95. Aberdare, Y Gwl.: halas i (3rd sing., fyclda), 30, 6,
1860, etc. In the Northern dialects -is, -ist occur from the
16tli century down. The explanation of them is uncertain,
but they are not due to any analogical neo-formation, since



the same change of ei into i oecnrs, in these very dialects, in
]ilurals of nonns like llygid (in Denbigh, Flint, and Meri-
oneth, Wliams ab Ithel),ifinc, bychin, etc, and in cymmint,
isio, iste, etc. (see my Beitr., 92). Davies reproves them
ruthlessly: ceris, cerist: summa3 imperitiie, poetice solum
fAifTjTL/cŵ' vel epwviK)<;.

Cf. Salesbury, N.T.: roist (?), f. -41. Add. MS. 14,986, 16th cont.:
rois, f. Yob; rois, f. 29 (ceres, f. 32a). Add. MS. 15,059, 17th cont.:
mi gollis, a gefis, f. 223. Add. MS. 31,056: a rois fy myd, f. 176: y
doist, f. 106; 15,005: oscefist, f. 136a. YrArw. (Pwllheli): miarois,
na welis i, 17, 7, 56; a ffen dois mi glywis, 31, 7, 56; ni welis i
mono chi, 13, 10, 56; na welist ti, 11, 12, 56; cefis, syrthis, 26,
2. 57; mi dtffris, mi ail dryehis arna ti, torist, 9, 4, 57. Cb.few T.:
mi gymris, p. 73; mi glowis, p. 49; glywis, p. 7; mi gymist dy,
p. 137 (cymmeraist): roist, p. 89; deydist, gelwist, etc.

19. The ending -ost (-os-t) is proper to buum, bm, and

was thence extended to daethum, gwnaethum, etc, which

follow the conjugation of buum. In old MSS. -ost is found

lIso in other verbs; cf. o buosty ema ty haythost, MS. A,

p. 71; ny cheuntoste, B. of Carm., 5; royssosti, 11. Gu\ L'h.,

p. 129. On the other hand these few anomalous verbs are in

later times assimilated to the ,s-Pret.; cf. wyddest ti,

Y Cyfaill Dyfyr, 1883 (Powys); ti wyddest, Scr. Cyrnru, ii,

pp. 48, 184; a fuest ti, i, p. 272, iii, p. 184; na fuest tithe,ii,

p. 423, etc Pthŷs, Rcr.Cclt., vi,p. 20, gives "bues-ti", or rather,

"bis-ti", as the colloquial brm in parts of Soutli Wales, and

compares Corn. fues, ves (Z., 2 p. 562). Since tlicse forms iu

-est are altogether wanting in older MSS., as far as I know,

I think it more probable tliat they represent a literary *buaist;

see 18. Williams ab Ithel gives Venedot. bum, Dimet.

buo, bues (lst sing.); Spurrell, Gramm., 3 189: Soulli- V r elsh

buais, bues, buo; Rowlands, Gramm., 2 248: buais, buaist.

Cf. Ser. C, mi fues, i, p. 411; iii, p. 103. Gwnais, dais, are

also given; even eisym; for in Davics's time, as he says,

cersym, caresym, archesym, ceusym, began to be formed, com-


binations of the s-Pret. with tlie endings of tlie perfect in
-um (later -yrn).

20. The ending -ast is of extremely rare occurrence, and
open to certain doubts. I have found doethast, Eev. Celt., vii,
p. 450; y deuthast, B. of Ecrg., Mab. Ger. ab Erbin (gwdast,
Mab., iii, p. 88 [Guest]?); pan vwryast ti, Hengwrt MSS., ii,
323; nadywedast, M. (Owen, Laics, p. 516); Salesbury, N. T.\
a ddaethast ti, f. 12b; canys beth a wyddas 1 ti, wreic, a ged-
wych di dy wr, neu beth a wyddos ti, wr . . . ., f. 250. As
gwybod and dyfod are each represented twice among these
few examples, the assumption of -ast as a real termination
gains somewhat in probability; the change of -ost and -ast
is either an analogical imitation or a real instance of the
interchange of o and a in the verbal terminations (-om, -ont)
in Welsh itself (buont: buant), and in Welsh compared with
Corn. and Bret. (Bret. queront: W. carant, etc.); see 26.

21. The 3rd sing. ends in -es, -is, -as, -wys (-ws). Tlie
first three of these endings are the reflex of the o-,j- and -
verbs, and the Part. Perf. Pass. in -ed, -id, -ad (*-eto-, etc.)
ordinarily corresponds to their formation. The regular rela-
tions, however, between'these two sets of forms and the 3rd
sing. Pres., the plural of the s-Aor. (-assom, -yssom, -sorn),
the old passive, certain infinitives, etc, are so often altered
by analogical neo-formations that it would require full collec-
tions from Welsh texts and a comparison with the other
Brythonic languages to ascertain the genuine formations.

Cf . a few f orms in -es: adoles, agores, annoges, anfones, arlioes,
bodes, canmoles, colles, cyfodes, cyffroes, cynghores, ynidangosses,

1 This orthography (wyddas ti) is very common with Salesbury. Cf.
ny wyddos ti, f. 155; ac a weles ti, f. 147; etc. In Hgt. MS. 2U2,
f. 1186, col. 2, line 22 (of the photograph in Y Cymmr., vii), gorugos
occurs (before a following vowel), an unique example, and not sumcieot,
I thii)k, to justify the assumption of the separate existence of -os at
that tirne.



dehogles, diffodes, dodes, esgores, etholes, ffoes, gosodes, gwrthores,
hoffes, holltes, llosces, paratoes, porthes, priodes, rhodes, sodes,
torres, troes, trosses, ymchweles; iu -is: edewis. erchis, ket\vis,
dyrcheuis, delis, kyuhelis, dechreuis, diengis, enwis, gelwis, genis,
gwerendewis, cyulenwis, nienegis, peris, seuis, tewis, trengis, trewis
(all from Ll. Gw. Rhydd.). Gwelas occurs at least 31 times in the
texts prnted from tliis MS., and 24 times in the Mab. (1887); also
welat, Ll. Gw., p. 206 (2); Mab., col. 734; but gweles exists 18 times
in L/. Gw. (pp. 302, 305, 309, written welais, pron. -es), andalsoin
otber MSS., e.g. t ILjt, MS. 59: weles, p. 417, viii, 5, 19; welas, p. 431
(v. Celt., vii); etc.; gwelas, gwelad are the commonly used forms.
Peir, peris, paryssei (Ll. Gu\, p. 113), or dieing, diengis, diengid,
dihagyssei (B. of Hcrg. , col. 164), form a regular series of the /- class.

22. The existence of the termination -wys (-ws) makes
the collection of the primitive forms more diffcult; for -ws,
which occurs also in the oldest Venedotian MSS., became at
a later date a characteristic of the Southern dialects, in which
it supplanted the genuine endings. Cf. deuenus, MS. A,
p. 1; a uarnus, p. 13; e lluydhaus, e delleghus, p. 50; talus,
pan Yihaus (gwrhau), p. 48; MS. B, pp. 473, 469; etc. It is
frequent in Ll. Gw. Bhydd., written -wys and -\vs (pro-
nounced -ws). Davies, Gramm., says: demetice -ws; E.
Lhuyd, A. Br., p. 239: S. W. -ws, -es, -ys (-is); L. Morris,
Add. MS. 14,934, f. 25?>; Pughe, Hughes, Williams, etc.;
T Traeth, iii, p. 11: S. W. rhows, cafas, rhanws, pryDwys,
cwnws, Efustws, etc. In Barddas, -wys and -wyd are exces-
sively common (gwelwys, cafwyd, etc). In modern texts
from Ebbw Vale, Monmouthshire, in P. G. 28, 29, -ws is often
used; in Y Bed., etc. In my opinion, as I stated in Obs. on
the Fron,, 07, -wys is shaped a'ter the model of -wyd; -es,
-is, -as, and -ed, -id, -ad, being exactly parallel in formatin;
and -wyd would seem to be an abstraction from bwyt, the
l'art. l'erf. Pass. of bod; possibly from *b(i)v-eto-. This erro-
neous abstraction was facilitated by the employmeut of this
bwyt toform passive participles of gwneuthwr, etc. (gwnaeth-
pwyt, etc). The 3rd sing. -oedd i'or -odd (see 10) offers


an analogy to this influence of the verb substantive on the
^erb. 1

23. In the plural we have three series of endings:
-assom, -yssom, -som; -assoch, etc.; and correspondingly in
the secondary s-Pret., -asswn, -ysswn, -swn, etc. The follow-
ing verbs refiect more or less constantly the old o-conjuga-
tion; cf. from texts:

B. of An.: nyt atcorsant, p. 96. B. of Tal.: digonsant, p. 127.
Tit. D 2: nỳ digonsant, f, 45a. B. of Ilerg.: a ducsynt, col. 165.
Ll. Gw.: ducssant, p. 203; niegys y kynierssom ni, p. 209;
a gymerssant, p. 225. Add. MS. 19,709: pan welsont 6y, f. 646.
Cleop. B 5: rỳwelsei, f. &a; n welssit (pass.) y chỳffelỳb, kl\vssei,
f. 10a. Dechreu, cafael (cawssom); and the verbs in -aw (infnitive):
ymadassant, B. of Herg., col. 595; val y hedewssynt, trawssey,
Ll. Gw., pp. 88, 254. Salesbury, iV. T.: gwelsont (often), cawsont,
f. 36; dechreysont, f. 18a; cymersant, f. 391a; clywsam, f. 179.
Gambold, Gramm., 1727, gives tawsom (tewi, s- pret. tewis), clyw-
son, gwrandawsom.

Notwithstanding these regular forms, which are further
proved to be old in many instances by the Part. Pret. Pass.
(see 31), we find a gymeryssant, Ll. Gia., p. 144; clywas-
soch, p. 77; nychlywyssynt, B. of Herg., col. 643; Sales-
bury, N. T.: cafesont, f. 159; ducesant, f. 179 (-esont,
f. 185); pan welesont, f. 208&; Mi/v. Arch., 2 p. 601: a wela-
sant (X. W., Hcm. Gruff. ab Cyn.); Add. MS. 15,056:
a welasant, f. 2a (O. Jones); also Barcldas, i, p. 236; etc.

24. It will be, I fear, impossible to define the exact limits
of -y-ssom, for no verb exists using exclusively -yssom, etc,

1 There occur, curiously enough, though in different texts, in the
same verb a fact which prevents my regarding them as undoubted
clerical errors forms of the 3rd sing. in -os: MS. Z (Owen), p. 526:
a gavos (1480); Salesbury, N. T.: ni chafos (f. 342a, Richard Davies);
Add. MS. 14,986, f. 15a: i kafos. I never found -os with any other
verb. Of course, if these forms really exist, they are the result of
momentary combinations of cafas and cdlodd cadd, for no othcr ortho-
graphies in these texts would justify our regarding s as a phonetic
orthography for dd (like s for th in MS. A).



and whilst in Micldle-Welsh texts -assom, etc, predominates,

the case is just the contrary in texts o' the lfith and 17th


Cf., e.g., B. of Herg.: dihagyssei, col. 161; ymchoelyssant, 166;
y kychwynyssont, col. 561; dodysson ni, col. 725. Add. MS.
19,709: disgynyssei, magyssei, f. 25a; karyssei, f. l9/>; yrnwa-
hanyssant, na phallyssam, lladyssn, eistedyssant, treulyssant, body-
ssant, etc. Ll. Gw. Rh.: pwy a royssei, pp. 122, 180 (roessei, p. 71;
roessant, p. 73; rodasswn, p. 80); dyrchauyssant, pp. 65, 145;
kanyssant, p. 65; ymgadarnhyssant, p. 50; ymrwymyssant, p. 84;
cyrchyssant, pp. 102, 131; ffoyssant, pp. 126, 128; cladyssant,
p. 117; gouynyssant, p. 137; ymgyweiryssant, p. 145; ymhoylys-
eant, pp. 139, 146; gorchymynyssant, p. 137; trigyssant, p. 146;
gwisgyssant, p. .138; archyssei, p. 120; paryssei, p. 118 (-ant,
p. 128); lladyssei, pp. 125, 128; dalyssei, p. 147; rydhayssei,
p. 131; rydywedyssei, p. 130; anuonyssei, p. 121; a detholyssant,
gordiwedyssant, chwardyssant, cussanyssei, ranyssant, priodyssei,
hwylyssant, etc.

25. In texts of the 16th and 17th centuries -yssom, etc.
(written -ysom,-esom,-isom,dialectal orthographies expressing
the different pronunciation of y) are the common endings.

Salesbury has, e.g., ymovynesei, roesei, roesont (Gicel. leuan: rois-
ont, f. 896/0; roysont, f. 38b; mi a rroysym, f. 375b); canesam (and
canasam), bwriesom, govynesoch, neidiesoch (marg. dawnsiesoch),
gwatworesont, a wnethesei, etc. Gu-el. I.: canysont, pan agoryssey,
f. 379/;; agorysei, ib. (agorassey, f. 882a; agorasay, f. 880o); etc.
Lewis Dwnn's Pedigrees: a sgrivenysant, p. 7; mynyswn, p. 9.
Lhjfr Achau (1604): a wledichison, a enillyssont, p. 60; cf. also
the Jlomilies (1606); Marchawg Cru-jdrad (17th cent.), etc.

At the same time, or a little later, -som, -soch, -sont
become the colloipiial ibrnis, and tliey are tlie only endings
used in the modern dialects. Cf. Add. MS. 14,921 (r.th
cent.): ymladdsont, f. 53a; lladson, f. 467 j hesides ni diodys-
son, f. 20ce; dodysson, f. 46a; ef a fynse and ef a fynysse,
ff. 226, 9a (kymersson, f. 46; kly\vson, lst pl., f. 13). Oann.
y Cymry, 1672: lladsont, p. 2(i; pan gotson, p. 32 (codi);
maethsont, p. 227; y gallsc, p. 139; ni haeddson, torsom,


p. 63, etc. Examples of -som, -socg, -sont and -swn, etc,
-sem, -sech, -sent (in Carnaiwonshire '-sam, etc, Sweet,
p. 444), are not necessary, as there are no exceptions. So
modern Welsh coincides in this point with the Breton and
Cornish languages, whilst the older language keeps up the
distinction between the three series. Carsom: caras followed
of course gwelsom: gwelas. Davies gives elswn, delswn,
gwnelswn, as Powysian forms. 1

26. -oin and -ont, corresponding to the Breton and Cornish
endings, are comparatively infrecpuent in Middle-Welsh,
but are the conimon modern Welsh endingo, a fact which
must, I thinh, be attributed not to the ancient interchange
of -om and -am, but to a transference of the vowel of the 2nd
plur. -och to these terniinations. I cannot make out the reasons
of the former duplication of endings, since we have also -ant,
Bret. -ont,in the Pres. Indic; -ach, Bret. -oc'h,in the compara-
tive, and interchanges of o and a in Welsh itself ( see my Beitr.,
55). Cf. B. of Herg.: rywelsom, col. 733; a dodyssom ni,
col. 725; a gassont, 742 (a gassant, 743); a glysont, 724;
kych6ynnyssont, 561. Lt. Gw. Rh.: ac am nas cawssont dic
uuant a thrist, p. 167; yd adawsont, p. 167. Both -om and
-am, -ont and -ant, occur also in the Pret. bm, aethuni,
daethum, etc; cf. buam, buant, Mab. (1887), often; a wdam
ni, col. 743; a 6dom ni, col. 742. Ll. Gw. Bh.: val y buont,
p. 106; a vuont, p. 162; uuont, p. 161; yd aethont, pp. 167,
] 73; a deuthont, p. 168. In later texts always, cf. Salesbury,
N. T.: gwyddom; ac ymaith yr aethon ein dau, March.
Crwyclr., p. 5; ni addoethon, p. 15; ni aethon, p. 140;

1 Sweet, Spolcen North Welsh, pp. 446, 448, tnentions the plur. rhodd-
son, rhoithon, rhoison (Sec. Pres. rhown, rht, rhy, rhown, rhoAvch,
rhn, rhoythan; s-Pret. rhois, rhoist, rhoddodd, rhth, rhs [rhys?]),
and danghothson (dangos); forms which the assumption of complex
analogical influence might render plausible, but which are too little
known to allow of a judgment on them.


y buont, p. 6; etc. Sometimes aw is written br o: ymnertha-
sawnt (Salesbury; etc.) pronounced o. 1

E. Passive.

27. E. Evans (Stud., G) first drew tlie attention of
pliilologists to the various old passives in -etor, -itor, -ator,
etc, which, though mentioned in the grammars of Pughe
and Williams ap Ithel, had been ignored by Zeuss and
Ebel. On edrychwyr and the other deponents see Rhŷs,
Rev. Gdt., vi, pp. 40-49, also in Y Traeh, 1S84.

Thefollowing forms occur: dgettaur, B. o/Carm., 7; B.ofTal.,
5(2); hymysgetor 30, lloscetar 5, golchettar, p. 16C, crybwyll-
etor (Evans); henhittor ldrn, B. of Carm., 17, megittor 17,
keissitor, p. 157(2); llemittyor, B. of Herg., p. 305 (Sk.); cwynitor,
gweinydiawr, gwelitor, clywitor, telitor (Ev.); traetbator,.B.q/"7 , (/.,
p. 131; molhator, ib.; kwuhỳator, B. of An., p. 86; gwelattor,
canhator (Ev.); brithottor, B. of Carm., 9.

Passives in -awr (Zeuss, 2 p. 529) are gnarwỳaur, B. of C, 17;
ergelar, ergelhar, B. of Tal., 5; dottar, gyrrar (a mettar am
dottar yn sawell ymgyrrar ymrygiar o la6), p. 136, gorff-
owyssar, p. 165, a uolhar, p. 165, dydyccar, p. 166(2); a emda-
flawr, na pharawr, B. of An.; etmychar, B. of Herg., p. 222; ni
chaffar, p. 235. Many other examples will be found in the poems
of the earlier incdi?eval bards.

Except in these old poems and others in the Book of Hcrgest
these forms occur, to my knowledge, only in Arwydon riju dyd
brawt (Ll. Gio., p. 274, gwelhitor), and in late Gwentian texts, as
Barddas, i (pwy arweinittor y treigl hwnnw? p. 244, ar a welittor,
p. 246, gwydditor, bythawr, p. 152, a deall a dichonawr arnaw,
p. 330), and the Thirteenth Book of the Welsh Laws, written in
1685 in Glamorganshire (-awr only): mal y gallawr ymgyunal,
p. 635, onis gwaretawr, p. 636, a wnelawr, p. 647, na phei heb
hyny y bythawr, p. 648, gallawr, p. 647, a ddotawr, p. 650, a
gatwawr, ]). (55, a gymcrawr, pp. 659, 677, barnawr, p. 674;
evidently in most cases modern forms (dodi, cymmerid, etc.)

1 Cf. B. of Hirg., col. 164: gedy daruat hyny (darfod); U. Gvo.:
drossawm, p. 212. D. S. Evans, Llythyr, 197, mentious that Dafydd
luuawr wrote -awu, -iawn, -awd, fur the plur. terminations -on, -ion, -od !


F. Part. Pret. Pass.

28. Phŷs, Rcv. Cclt,, vi (Some Welsh Deponents, 9), tries
to explain the now existing use of these forms in -et, -it, -at,
-wyt to denote the perfect passive, by assuming them to
have been originally deponent participles; the passive form
of bod, giving the passive meaning, having been omitted.
This theory, of course, does not affect their passive character
in point of forrnation, which has in part been very largely
altered by the subsequent effects of analogy, and corresponds
in a certain degree to that of the s-aorist. Lls and gwŷs
(Zeuss, 2 p. 531) are remains of the old participles in -to- (s from
*d-t). Williams (Dosp. Edeym, 717) gives lls as a Gwentian
form. It is very frequent in the Middle-Welsh MSS., but
lladded and lladdwyd are not wanting (cf. lladyt, B. of
Hcrg., cols. 685, 841, 842(2), etc). Gwŷs occurs in modern
Gwentian dialects; cf. wys neb i 'ble, Y Tyw. ct'r G., ii,
p. 241, etc. Gwyddys is a combination of *gwydded and
gwys; cf. ni wyddys, Hgt. MSS., ii, p. 327; Davies, Gramm.,
etc. A single instance of such a neo-formation if not a
clerical error is also lladass, B. of Hcrg., col. 841 (besides
llas and lladyt, ib.). 1

29. We find forms in -s compounded with bwyt, -pwyt,
formed from nearly all the verbs which append the termina-

1 If we notice words like gorddiwedd and gorddiwes, etc. (see 32),
it is difficult not to assume a siniilar connection between lladdu and
lleassu; lleas (leturn, caedes, Davies, Dict.) is not infrequent in Middle-
Welsh; cf. cyn Ueas, B. of Tal., 41; o leas cledeu y teruynir, Ll. CJir.,
p. 283; lleas, MS. Tit. D22, f. 13a; rac drycket gennyf gelet lleassu
(to becoine slain) gas kyn decket a thi; p6y am lleassei i heb y
peredur, B. of Uerg., col. 679; pan allo lleassu pab uelly, col. 680;
pan yinleasso, Didr. Casgl, p. 248 (Oderic's Travels): if he kills him-
self. In Myv. Arch., 2 p. llb (Gododin), gwnelut leadut llosgut is printed.
and ladut is given from the other MSS. Leadut is an unique form.
Probably the old lleassut was replaced by the later lladdut; hence the
conf usion of the transcriber.


tions -som, etc, and not -assom, -yssom, to their "Welsh
stems to form the s-aorist; e.g., dechreusom dechreuspwy t.
Dywespwyt occurs more frequently than any other of them,
and is said by Davies (Gramm.,\). 197) and Williams, etc, to be
a Dimetian peculiarity. Notwithstanding this, which the
later MSS. indeed tend to affirm, it is also found in MSS. of
the Venedotian Laios, viz. MS. Tit. D2 (=B), a dỳwesput,
ff. Ob, 42; rdỳwesput, f. 48; MS. Galig. A3 { = 0),
redỳwespwỳt, f, 1786, But it is incomparably more frequent
in S. W. MSS.

Cf. dỳwespyt, Owen, p. 172 (3 MSS.), p. 212 (7 MSS.), pp. 188,
253, etc.; B. of Eerg.: dywespyt, col. 317; Add. MS. 19,709,
f. 47; Cleop. B5, f. 227 2; Tit. 1)22, ff. 7, 15a, etc.; Add. MS.
22,356 (=S): deesbyd, p. 195; deesbyd, p. 550; desbyd, p. 593
(dyespyt, p. 212). (Sal., N. T: dywetpwyt, f. 4a; dywedwyt,
dwetpwyd, f. 380a; doytbwyd, f. 3306; doytpwyt, f: 33); Add.
MS. 14,921 (16th cent.): dwesbwyd, f. 38; wespwyd, f. 23, etc.
(7); wesbwyd, f. 34; a ddwedsbwyd (sic), f. 41a; Medd. Myddfai:
a ddywespwyd, 94, 97, 98, etc.; Barddas, i: wespwyd, p. 208, etc.

30. In Harl. MS. 958 (=T), f. 40, is written: kht
ac y dwedaspỳt 6r. A clerical error may certainly be
surmised here, the transcriber having possibly intendcd at
irst to write dwedas sam (we have said), and omitted to
erase -as- after writing -p6t. But if we recall "a dywed-
adoed",MS. Cleop B5, f. 247 2 Dares Phrygius), and "wal y
dawedadoedit idaw", Ll. Gio. Bh., p. 269, we may with confi-
dence hold dywedas- to be a combination of dywedad- and
dywes-; cf. gwyddys. Similar neo-formations are mynassuedd,
planassoedd, rodassoedd. Cf. y mynassoedd Cli. ymhoelut oe
nerthu (Ch. would have returned to help him),Z/. Gw., p. 108;
y gwielin a blanasoed voesen yno, p. 246; rodassoed, MS. B of
Brut y Tywys., p. 290 = roessoed in the B. of Herg. I am
unable to accede to the obviously more simple opinion that
mynassoedd: mynasswn (second s-aorist) is an imitation ff
aethwn: aethoedd, doethwn: doethoedd, etc, owing to the


numerous forms which exist in -ad-oedd, -yd-oedd, evidently
containing participles. On these see E. Evans, Stud., 20,
and Ehŷs, l. c. Cf. MS. Gleop. B5: archadoed, f. 230 1; Add.
MS. 19,709: ny orffoyssys corineus or ruthur hono yny oed
kainn6yaf ỳ elynyon yn anafus ar ny ladadoed onadunt,
f. 15; Ll. Gw. Bh.: mal y barnadoed idaw, p. 270; nyni
ganadoed yna, p. 263; y haffadoed, p. 265; rynodydoed, B.
of Herg., col. 841; dysgydoed, Ll. Gw., p. 135; ganydoed, pp.
135, 142, 154; Y S. Gr., p. 377; Add. MS. 19,709, f. m> etc.

31. Forms similar to dywespwyd occur further in:

Glywed: a glywspwyt, Y S. Gr., p. 216; ny chlywyspwyt,
Yst. Gwl.
Ieuan Fencl, p. 327.

Dechreu: dechreuspwyt, Y S. Gr., p. 29 .

Bhoi, rhoddi: rossoedit, Brut y Tyw., MS. B (Dimet.),
p. 192; y rossoed, Y S. Gr., p. 236; a rassoedynt, p. 399;
a roespwyt, pp. 238, 354; rodassoed, roessoed, see 30;
Salesbury, N T.: a roed, marg. roespwyt, f. 237; a roespwyt,
marg. roddet, f. 227; rhoespwyt, f. 293, B. D., f. 3176;
rroyspwydd (sic), Gwel. leuan, f. 381.

Cafacl: 1. Y S. Gr. (a MS. abounding in these forma-
tions): a gawssoedd, pp. 303, 382; mi a gawssoedwn, pp.
231, 278, 322; pei cawssoedut, p. 247; hawssoedyat, p. 406;
nych., p. 208; c, pp. 306, 313; Ll. Giv. Bh.: hawssodyat, p.
224 deg. hawssoydat?); y cawssoedat, ib.

2. Pei cassoedyat, Y S. Gr., p. 297; ny chassoed hi, p. 412;
pei cassoed} T nt, p. 429.

3. Didr. Gasgl., f. 398: ac ymgaru yn va6r awnaethant yn
y byt h6nn try weithredu bei katlioedynt gyfle a meir6
vuant heb gyffessn y p6nk I1611116 . . . .; Sal., N. T.: pa vodd
y cawseief, marg. cathoddei, f. 148; Y Drych Ghrist.: 1585,
a gathoedhei, f. 30.

Cathoeddwn is formed after the model of gwnathoeddwn,
etc. In the modern dialect of Carnaiwonshire this imitation,
caused there probably by the s-aorist of aeth, doeth, etc. (s


coinciding in appearance with cs, / had), took place on a
large scale. Sweet, p. 450, gives, Sec. Pres.: kawn, kaythat,
kv, pl. kaythan, -ach, -an; s-Aor.: kefs, ks, kst, kafodil,
kdd, kth, pl. kaython and koyson, etc. (like awn, ewn,
oythwn?); t, y, pl. oythan, n (?), oythach, -an; ais, s, oist,
st, th, pl. oython, oyson, etc.

4. Y kaffadoed, Ll. Gw. Rh., p. 265.

5. Cespwyd, Bardd., i, p. 268; Evans, Llythyr, 162.
Gweled: gwelspwyt, Sal., N. T., f. 259; y welspwyd,

f. 386.

Gymeryd: y gymerspwyd, Sal., N T.; Gwel.Ieuan, f. 386.

Gwneuthur: gwnespwyd, Barddas, i, p. 250; gwnaes-
pwyd, ib.

32. Gwes-, dywes- in dywespwyt cannot be an old par-
ticiple in -to-, since t(*v<t-) + t is known to become tk in
Welsh. So it must be the result of *vet- + s, like the s-
future of gwared: gwared, Perf. gwarawd, s-Fut. gwares
(*gwo-ret-s-), Rhŷs, Bev. Cclt., vi; so dywed,dywawd (dywod,
dywad), *dywes. These forms in s of dywed do not seem to
have been all lost, for it is just of this verb, and, as far as I
know, of no other in the same texts, that s-aorists like
dywessont occur. It is perhaps not practicable to regard
them all as clerical errors, for why should clerical errors
occur in the same word in different texts, and in that word
alone? At the best one may look upon the more recent of
them as recent formations caused by dywespwyt itself.

I have found, B. of Herg.: y dywessont, col. 804; L. G>r. lh.
'Y Cfroglth): Tony ychyna deueist dyhun. heb y yrenhines. bot y
<,'\veithredoed yn betruster ydyweist. heb y iudas, p. 262 (a corrupt
pissage); MS. 3 (Owen, p. 515): a ddewast di; MS. /1 (<!>., p.
768): dywedeist and dy weiat; Salesbury, N.T. .- dywesont. f. 1706;
ry ddywesei, f. 17-ia (y dywetsei ef, a ddywetsont, ff. i'2b, Sb; a
ddywetsant, f. 46/>; ddywedsant, f. 7l)b); ctc.

Dyweist: dywesam, probably like careist: caryssam, is
perhapa due to a wrong separatiou (dyw-esam); dywessom is


not affectecl by the existence or non-existence of dyweist.
Perhaps the coexisteuce of rhoddi and rhoi (but cf. also Bret.
reif, Corn. ry) is due to the formation of an old s-aorist:
*rhossom (lst pl.), or a -Part. *rhos, the existence of which
the above-quoted rossoedit, rossoed, tend to establish. In
*rhossom the d (of the root *d) perished in its encounter
with s, and hence rho-, like gwel- of gwelsom, seems to have
been wrongly abstracted and transferred to other tenses. I
suppose arhosi and arhoi, also 3rd sing. erys and ery, etc. (Ir.
arus), have been differentiated by the same analogical in-
fiuence of an erroneous separation in an s-form.

33. As far as this point phonetic laws empower us to
go. For what reason the s-form dywes- was introduced in
the place of an old Part. Perf. Pass. I cannot ascertain. It
is, however, clear that the other s-forms (clywspwyt, etc.) are
imitations of dywespwyt, caused by dy wessom: dywespwyt.
All these verbs have commonly -som, and not -assom, -yssom,
except rhoddi, rhoi (see 23). Ehoespwyd, cespwyd, gwnes-
pwyd, have to all evidence -es- directly taken from dywes-;
while rhossoedd, cawssoedd, are imitations of the principle
of dywes-. Rhoespwyd, cespwyd, have advanced a degree
further, and imitate it even literally. 1

34. As to the single participle in -ed, -id, -ad and (-eid),
-wyd, the first three must be separated from -wyd, which is a
South-Welsh termination, replacing more or less all others in
certain texts.

Cf. ỳ rodat, Clcop. B5, f. 161a (ordinarily rhodded); y rrowd,
Sal., N. T, f . 382Z, (Huet) (Add. MS. 14,973, f. 79, trowd); rhodd-
wys, Cann. y C, p. 404, Trioedd Barddas in Poems Lyr. and J'ust.,
ii, p. 238, etc. L. Morris, Add. MS. 14,945, f. 273a, transcript of
an old MS.: pan wneithbwit, ac distrwt, ỳ gwisgwt (lej. wt).
In modern S.W.: rhow'd, aw'd, caw'd, gwnaw'd, clow'd (Y Tratth-

1 Rhoss, however, may be an old Part., and rhoes- a doublet of
rhoddes-, combined of *rhos and rhodded, like gwyddys.



odydd, iii, p. 11), clywspwt, bwriwd, clwyfwd; Y Bed. 10, p. 8,
ctc. -wt is written far less ofteu than -ws for -wys, since its con-
nection with the Verb. Subst. was still more or less transparent.
For a hypothesis on -wys and wyt, see 22.

G. Infinitive.

35. The infinitives in -o, Middle-Welsh, -aw, represent ah
older -*m-, like llaw (hand) from *plm. Llaw: unllofyauc
(Mab.), llofrndd, rbaw -iau and rhofiau (infin. rhawio, rhofio),
crawen and crofen (crust) are doublets of a kind which must
also be expected in the infinitive, and there are indeed traces of
infinitives in -of. Cf. gwallaw and gwallawf, findere, Davies,
Bict.; gwallo, gwallof, gwallofaint, Spurrell, Dict. Gwallot,
gwallofaint, contain gwallof (/ kept in the inlaut) + -i, -aint.
So probably also cwyno and cwynof-ain, to wail; cwynof-aint,
wailing; wylo. to weep; wylof-us, wylof-aeth, ' wylof (Sp.),
weeping. -if (frequent as -im in the old Brythonic glosses)
occurs (see Rev. Cclt., ii, p. 135) in the Mcb. Eilhwch ac Olwen;
cf. ae dodi ar eu hol. ac aruoll aoruc bedwyr ae odif ynteu,
col. 825; kymryt aoruc ynten yreilllechwae... ae odif areu
liol, ib.; and tliree other times, ib. I have not found it else-
where in texts. In gweinif-iad: gweini (Sp. has also gwenif,
ministration),/is kept in the inlaut.

36. The intnitive of the verbs in -aaf, *-agaf, ends in
-iiu, but also sometimes in -a. Davies, Gramm., gives bwyta
and llewa as South-Welsh forms; also Hughes, 18l ) i ) , p. 32:
S.-W. bytta; Y Gwyliedydd, 1828: S.-W. para, gwella
(=}iailiu, gwellhu). Bwyta (and byta; see my Beitr.,
103) occur, however, in all dialects; cf. buyta, MS. A,
pp. 15, 21 , 22, 23, 24, 3,0, etc.; Sal., N. T.: bwyta, f'. 40, 59a;
Add. MS. 14,913 (lGth ccnt., S.-W.): bytta, f. 84a; Sweet,
SpoJcenNorth-Welsh, p. 423: byta; Yr Arw.: i futa, 20, 1, 59;
uii futa i, Cb.few. T.; byta, Seren Cymru, i, p. 373, iii, 447.
This verb offers certain other particularities too. Bwytavss-
awch, y bwytawssant, Y S. Gr., 68, may be forms not to



be relied pn. In Ll. Gw. Rh., p. 265, wynteu ae kymerawss-
nthwy is printed, caused, if not an error, by the coexist-
ence of gadawssant, and, e.g., a adassant, MS. A, p. 1, MS. G,
p. 105; gadassant, B. of Herg., col. 705. The opt;itiv<>
bwytafwyf occurs in Salesbury's N. T, f. 124 (twice bwyta-
vwyf), and in Add. MS. 14,986, f. 336: ond kyn i bwytavwy
ddim bara. Davies, Gramm., quotes bwytathwyf (corrupte),
following, in his opinion, rhothwyf for rhoddwyf. This neo-
formation (if th, dd, f are not simply exchanged, as they are
in many other words) seems indeed to have sprung from an
erroneous abstraction of -thwyf in rhothwyf, since the other
forms of rhoddi, rhoi are partly used without dd (rhoddaf, rhoaf,
rhf, etc). In grammars (Gambold, Eowlands, etc) the 3rd
sing. bwytty,G.,bwwty, R.,pery (of parha, parhu),E., are given,
like gwrendy, gedy, tery, eddy, of gwrando, gado, etc. (Middle-
Welsh gwerendeu, gedeu, tereu, etc). 1 So bwyta partly
follows the analogy of the verbs in -aw, -o. Spurrell has
ciniaw, cinio, ciniawau, dinner; ciniawa, to dine. Many infi-
nitives in -a are formed from substantives, and express a
collection of the thing which the substantive denotes (" verba
colligendi", Davies): cneua, afaleua, ytta, gwiala, coetta,
cawsa, ceiniocca. They contain real plurals, like afaleu,
cneu, gwial, etc.

37. E.Evans (Stud., 27) first noticed the Infin. cadwY-l
(besides cadw); an imitation of dywedyd, dychwelyd, in his
opinion. Dywedwyd, on the other hand, occurs not infre-
quently in very old texts for dywedyd; also cyscwyd; B. of
Tal.: a ddostti peth 6yt pan vych yn kysc6yt.

Cf. MS. A: deueduyt. pp. 43, 46, 52; deweduyt, p. 66; deueduit,
p. 63; deveduit, p. 70; deueduyd, p. 52; dewetduyd, p. 71; dewetuyt,
p. 70; dewedyyt, p. 70. MS. Tit. D2 (=B): dweduet, ff. 18, 1%;

1 Cf. Lew. Gl. Cothi (ed. 1837): chwery, tery, p. 41; ni thery, e dery,
p. 102; pery, p. 134 (ŵ, a dau, p. 31; D. ab Gw.: ni thawaf, o thau, p.
253, froui tewi).




deweduet, f. 20b; dỳweduỳt, f. 63. Ccrfig. A13 (= C): dewedwt,
f. 152; dwedwỳt, f. 179. Add. MS. 14,931 (= E): dỳwedut,
f. V2b. Hctrl. MS. 958 (= T, Dimet.): dwedỳt, f. 20a. Add.MS.
14,903, a 17th ceut. MS. of Br. y Brenh.: doydyd, f. 20a, besides
the old dỳwedwyt, ib.

Also in old derivatives: Pa dywedydat oed h6nn6, B. of Herg.,
col. 740; dywedwydyat, a babbler, Y S. Gr., p. 171; ar eu dywe-
dwydyat, talkativeness, p. 179; Add. MS. 14,912, f. 496, dywe-
dwydal. In the translation of Dares PArygius, in the old MS.
Cleop. B5, dywedwydawl is written (ff. 228al, 229a 1), where in
the later MS., Jes. Coll., 141, the modern dywediadawl has been
introduced (f. 19a). See also Pughe's Dict. s. v.

So the older yachwyawdr (c.g., iachwyawdyr, Y S. Gr., pp.
240, 260, 268; iachyadyr, Tit. D22, f. 185a; yachwyawl,
Ll. Gw. Rh., p. 264) is now iachawdwr (saviour). I carmot
say whether dywedwyd is formed after the model of cad-
wyd, though the pronunciation of cadw as cad seems to support
this explanation, for cad, cadwyd, 3rd sing. ceid (pron.),
dyweid might easily have given rise to the wrong notion of
an ending -wyd (-vyd).

38. Tn later Welsh -ud, -yd, -id tend largely to replace
the infinitives in -u, -i. The verbs dywedyd and ynich-
welyd (in Middle-Welsh ending in -ut), which occur so
very frequently, were the models of these neo-formations.
Salesbury, N. T., has, c.g., gwneuthur (gwnauthur, f. lOoa;
gwnewthur, f. 240; gwneuthr, at y C.; gwneithr, f. 377/',
Huet); gwneuthu, f. 91, gwneuthy, f. 215&; and gwneythyd,
gwneuthy'd. He prints also wneythy'r, d'anvon, etc. In the
modern dialects gwneyd is the common form: cf. Yr Arw.:
gneud, gnyud, 17, 7, 56; Cardigansh.: gneid, Y Cymmr., v, p.
122; Aberdare: gneud, gneithir, Y Gwcithiiur; Ebbw Vale:
wedi gnd nyna (= hyn yna), P. C., 28 (ib.: gwd, to say).
Gneyd is an imitation of dyweyd, gweyd, used iu the same
diiilects; see 42.

J. 1). Rhys, 1595, givesin his Grammar, p. 128, mynnud. )V Arw.'.
carud, 18, 12, 56, pcrud, tawlud, oanud, yn sgwenud (ysgrifonu),



prydDawnud, 17, 7, 56, collid, 12, 2, 57. Cab.few. T.: gwerthud,
heuddud, penderfynud, gyrud, medrud, tynud, sychud, gallud
cnoid, cospid, collid, troid, rhoid, etc. Cwrddyd (e.g., V Gwron
Cymr., Caerfyrddin, 20, 5, 1852) for cwrdd is reproved, Y Gwlad-
garwr, 14, 5, 1858; etc.

39. -an, -ian is a very frequent moclern termination,
principally used in words denoting a repeated noise, and still
living, as it is the ending of many English loan-words.

Cf. Dimet. mcian,ppian (topeep), Powel; Cann.yCymry, 1672:
chwarian, p. 244; beggian, p. 128; begian (marg. ceisio), p. 61;
moccian (marg. gwatwor), p. 154; bribian, brwylian, etc. Seren
Cymru: yn clebran byth a hefyd am ryw gyfrinion sy 'da nhwy, iii,
p. 22 (clebar, clebren, Spurrell; IV Arw.: ac yn clibir di glebar yn
y riaith Sasnag, 17, 7, 56); llolian and llolio, hwtian and hwtio,
Sp., Cb.few. T.\ yn clecian ac yn crecian, p. 46; screchian, p. 120
(to screech). Y Genedl Gymr.: yn brygawthian, 6, 5, 1885, p. 7 3 .

40. Some minor dialectal differences are: dringad for
dringo, quoted in Y Traeth., 1870, p. 412, from Williams
Pant y Celyn; cf. ei dhringad, marg. dringo, Hom., 1606, iii, p.
217. Ib.: damsiad for damsang or mathru (to tread, to
trample), used by the same author. 1 L. Morris (Add. MS.
14,944) mentions nadu from Anglesey, nadel from Merioneth-
shire (na-adael). Hely and hela, daly and dala, do not conie
under this head, since their difference is a phonetic one, the
result of a different treatment of the group *lg, as in boly,
bola, eiry, eira, etc. Hla, to drive, to send, to spend, is S.-W.;
cf. E. Lhuyd, A. Br.: S.-W. hel, hela, to send; L. Morris, Add.

1 As to these words and their synonyms, cf. seingat, a trampling,
B. nf Au., pp. 85, 72; maessing, myssaing: sathru, G. Lleyn's Vocab.
Davies, Dict., s. v. calco: sathru, mathru, mysseing, sengi, troedio.
Sp. has maessyng, totrample about, and mysangu, to trample, the latter
for *ymsangu, hke mysygan f., soft expression, and sygan, whisper,
mutter. Homilies, 1606: a sathrodd, marg. ddansiolodd, iii, p. 121; y
methrer hwy dan draed, marg. dansielyr, i, p. 108; mathru, marg.
ddansial, iii, p. 272; ymsang, marg. ymwasc, iii, p. 261. Cann. y C,
1672: damsing, marg. sathra, p. 86; 'n damsing, marg. sathru, p. 198;
i'w ddamsian, marg. i gerdded trosto, p. 432. Llyfr y Resol.: N.-W.
fathru(i6.: also folestu, etc.) S.-W. damsang. Hughes, 1822: S.-W.
damsang N.-W. sathru. Y Gwyl, 1828: S.-W. sengyd N.-W. satliru.

G 2


MS. 14,923, f. 134: S.-W. hala, to send, to earry N.-W.
gyrru, carrio; YGwyl, 1828: S.-W.hla; 7 Traeth., iii: S.-W.
hala lii i'r lan = danfon lii i fynu; wedi hela triugain
mlwydd = w. treulio neu fyw t. m.; halas i, Y Gwyl. (Aher-
dare); etc. In Y Traeth., iii, p. 10: Dimet. meithrynu (mei-
thrin), barni (barnu), costi (costio), rhosti (rhostio); D. S.
Evans, Llythyr., 154: N.-W. peri S.-W. peru; N.-W. cymylu ■
S.-W. cymylo; N.-W. gwalu S.-W. gwalo; etc.

H. On the Verbs Dywedyd and Ymchwelyd.

41. These two verbs do not belong to the so-callcd
irregular class, but they exhibit in difereut periods and in
different dialects such a variety of external variations as to
justify mydwelling further upon thein. They present such
striking similarities to each other that copious e&amples are
necessary to show that these similarities are, in all probabi-
lity, not the result of mutual influence. The uncompounded
gwedyd is the common South-Welsh form. Cf. MS. TiL,IY22:
ar hii a wetto, f. 2 (Dimet.); Add. MS. 14,921 (16th cent.):
ny wedaf, f. 52&; hwy wedan, 'i'. 19, lb, 42, 43b, 5%, etc.
Hughes, 1822: South-Welsh gweyd, gwedyd. Seren Cymru:
gweyd, i, p. 192; wedest ti, p. 232; wedsoch chi, ib. (Car-
marthenshire). Y Gweithiwr: Imp. gwed, Inf. gweyd (Aber-
dare). Y Tyw. a'r G.: i weta chi, i, p. 94; gwetwch, p. 93;
ond wetas I, p. 95; Inf. gweyd, gwed. YGt ninen, iii, \k 19:
East Glam. i wd = North-Welsh i ddweyd. Y Bed.: wetaf,
'jiiu wetyd, gwed; gweyd (Monmouths"hire). P. C.,28: wedws,
gwd (Ebbw Vale); etc. 1 Dywed- is also written doẅed-

1 Further illustrations of different prepositions used in dialects are:
South-Welsh dillwng = gollwng, Davies, Dct. North-Welsh dobr =
gwobr, Han. y Ffydd, index. South-Welsb dyrru, to drive = gyrru,
Daries, Dict.; cf. ac achelary a dyrrad ffo;ir y llu h6n6, Jl. of
Herg.. f<>\. i' (in tliis MS. of the Dares Phrygiiis, but not in tlie other
four known to me, achelarwy is commonly written for achel, echel,
A.cbilles); ymddwyn, to conceive with young,in Powys dymddwyn, ib.\


(see niy Beitr., 40); and, owing to the loss of pre-tonic syl-
lables, dwed-; cf. Add. MS. 22,356 (= S): 3rd sing. deeud,
p. 591; deid, ib.; hyd y detto, p. 590; megis y dedir,
p. 593; desbyd, deesbyd (see 29); Add. MS. 14,912:
dwespwyt, f. 41a; dwesbwyt, f. 34&; dwetpwyt, f. 41, 42.
Salesbury, N. T.: ni dowot; dowait, dowaid, f. 3276. Athr.
Grist: doUaid,p.46; dowaid, Add. MS. 31,058,f.69a; dowedyd,
Add. MS. 15,058, f. 59 (17th cent.); etc.

42. Froni the 16th century downwards, dyweidyd, dwei-
dyd, deidyd, and doedyd (doydyd) occur frequently in books
and MSS.

Cf. Salesbury, N T: besides'dywedysont, f. 36; dywetsit, f. 4a;
dywedesit, f. 4&; dywetpwyt, f. 4.; dywedwyt, f. 46; dyvawt,
f. 149; dyvot, f. 2226; dowot, dowaid, etc; y doedaf, f. lb; mal
i doydais, at y C, doytpwyt, f. 30; Inf. doedit, f. 66; doydyd
doedyt. R. Davies: doydaf, f . 3126; a ddoytont, f. 31 1; a ddoyd
asont, f. 3596; a racddoydasant, f. S61b; y doytbwyd, Inf
doydid, f. 328a.
Huet: y ddwedasont, f. 3806; y ddwetpwyt
f. 3796; y ddwed, f. 3766; etc. Athr. Grist.: a doedasom ni
p. 14; doydyd, p. (3). Y Drych Christ.: doydais, f. 156; doyd-
iadeu r saint, Cl; dau dhoyded, Aja (also in Sal., N. T.: yn doyded,
f. 3606, R. D.); megis i dwed, Aij; o dweid, Bijb; u doedud pro
dowedyd et dwedud pro dywedyd", J. D. Rhys, Gramm., 1595,
. p. 128.

Add. MSS. 14,986 (16th cent.): dweydyd, f. 13a; dywevdwch,
f . 14; yn dywevdvd, f. 14; dywevdyd, f. 17; in another hand:
ar a ddywoydo wrthyd, f. 59a. Add. MS. 14,979 (Life of Petrus);
doydud, dowaid, often; dyweydyt, f. 157; a ddoydassont, f. 160&.
Add. MS. 14,913 (South-Welsh, lth cent.): val y dyweid, f. 16;
addweid, f. 15; doetbwyd, f. 8. Add. MS. 14,898: doidaist.
f. 72; mi a ddoyda, f. 406; doydyd, f. 416 (another hand). Add.
MS. 14,989 (17th cent., prose): yn dicoeydyd, ff. 87, 906; and
doydyd, ord sing. dowed, rhagddoedwr, f. 876 (in the same text
dallt and deyallt, clowed, edrch, sevll, cylfyddod, diwioldeb,
etc). Add. MS. 14,987 (17th cent., Araetli y Trwstan): doudud,
f. 816. Add. MS. 31,057: doydwn, f. 156; na ddowaid, f. 15;
a ddowaid, f. 156.

South-Welsh goddef (frequently in the Hom., 1606) = dioddef, Hughes.
1822; South-Welsh dyro = rho (give), Davies, Gramm., frequently
written doro, e.g., Add. MS. 14,912, f. 30; 15,049, ff. 3a, 36, 46, 5, etc


Add. MS. 31,060: ganddedyd, f. 2l7a. Add. MS. 15059: dwyd-
wch, f. 222b (North-Welsh popular lauguage, 18th ceut.). Add.
MS. 31,056: gwrandewch arnai y dwydyd fy rneigl om ifiengtid,
mi ddoudaf i chwi, etc, f. lla; deudyd, f. 18&; a ddeuden, ib.

m i ddoeda itti, f. 10b a ddoede, f. 34a; doyde, f. 199a; mi
ddweudwn, f. 20a. Add. MS. 14,97-4 (17th cent,): yn doeyd,
f. 78b; da niedrid di yn rwydd ddoevd kelwvdd, f. 79. Hope,
Cyf. Vr Cym.ro, 1765: a ddwed, p. vii; gwir a ddeidi di, p. 12; ui
ddweyde, p. 22; dweud, p, vii; dweudỳd, p. 32.

Doedyd is of most frequent occurrence in the works of
Griffith Eoberts, and in niany MSS. in prose and verse of
the lGth and 17th cent., some of which bear characteristics
of Northern dialects, as miewn for mewn, etc.

In modern dialects: Carnarvonshire: Infin. dayd; 3rd sing.
dwd, dfyd, doydith; s-Aor. plur. daydson and dwedson; Imp.
dwad, dayd, plur. dwedwch, daydwch, Sweet, p. 449. IV Arw.:
mi dyuda i chi, 20, 1, 59; dyudis i, 12, 5, 57; mi ddeydis, 17, 7, 56;
Iuf. dyud, ib.; dywudyd, 19, 11, 56. Merionethshire, Caban
f'ew. T.: deydwch, p. 19; wal y deydodd hithe, p. 36; yr hyn a
dwedodd, p. 36; deydist, pp. 7, 109; deydsoch, p. 475; Inf. deyd,
p. 56; deydyd, p. 7; deydis, Yr Ainserau, 17, 12, 1846 (Hen Ffar-
mwr). In the South-Welsh dialects gwedyd is used, on which see

43. Ymchwelyd: ymchoelyd, ymhoelyd apparently

represent the same phonetic- changes as dywedyd, dwedyd:

doedyd. This external similarity, however, is not sufficient,

and the time and dialect of hoth forms must be considered.

Ymhoelyd is said by Powel (annotations to MS. Tit., D22,

f. 15, Y Gymmr., iv) to be a Dimetian form; 'mhoylu teisen,

'mhoylu gwair, llafur (South-Welsh for yd), etc. It occurs

frecpiently in Middle-Welsh MSS., forms of which are:

MS. A: or pan emcuelo, p. 157; nyt hcmchuel, p. 159; kanyd
emchel, p. 46; cny emchelo, p. 392 (ib., chewraur, p. 68 = chwef-
rawr). MS. C, p. 157: emchwelhoent (D, B, K, yuichoelont).
!\1S. /,', f. 6O<0: pan einchuelhoent, the usual form in this MS.
MS. E: nỳ iuchwelo, p. 266; etc

B. of Carm.: mchueli, l.">; />'.
<>f Tal.: ymchflcnt y perth
gled, ynichoelant, 5 (in /./. '.'//. Rh.: ac amhaelawd ar processiwn
yr eglwya, p. L96)


MS. W (Gwentian Code, Cleop., Al4): hỳn ymhoelo, f. 60a;
ỳd mhoelir, f. 97, etc. MS. 5 (Add. MS. 22,356): yd ymhoilyr,
f. G46; ymchoylyt, f. 166; nyd ymhelant hynteu, f. 166; yd
ymchoylr, f. 486.
Add. MS. 19,709: ymchoelut, f. 226; ymhoelut
f. 10fl.

In the Mbinogion and Triads, B. of Herg. (Oxford, 1887), I
found at least 39 times ymchoelut, -yt (Infin.), -ad, -es, etc;
once ymhoelad, col. 672, and probably never ymchwelut.

In the parts edited by R. Williams f rom Lhjfr. Gwyn Rhydderch: in
Charhmagns Yoyage, only ymchoelut (p. 2, -es, 4), and ymhoelut,
p. 17, occur. In Turpin's Chronicle, ymchwelut, -ws, -assant, etc,
occur 17 times; ymchoelut. etc, 3 times (ymchoel, p. 87; ymchoylei,
p. 46; -wys, p. 44); ymhoelut, 22 times (ymhoylut, p. 43; ymhoyl,
p. 82, etc). In Bown o Hamtwn: ymchwelut,p. 127; ymchoelawd,
p. 188; ymhoelut (ymhoyly, 2nd sing., p. 138; ymoel, p. 138), 30
titnes. In Purdan Padric: ymchwelut, etc, p. 192(2); amhaelaud,
p. 196; ymhoelut, 10 times. In Yst. Giulat Ieuan Vend.: ymhoelyt,
p. 328, -u, 2nd sing., p. 328; ymhoel, pp. 328, 334 (Select. from
Hgt. MSS., vol. ii).

Salesbury, N. T: ymchwelyt, f. oa; ymchoelyt, f. 1716; nad
ymchoelent at, marg. y dychwelasant, f. 36; ydd ymchoelodd,
f. 266, etc. In later MSS. ymchwelyd usually occurs, also in
those which contain regularly doedyd.

44. Like dyweidyd (see 42), ymchweilyd occurs very
seldom. I found it in only two MSS., viz., in Cleop., B5
[3rd sing. ymchweil, f. 686; Inf. dchwelut, f. 1536; mch-
elut, f. 1386; dchelut adref, f. 152; Imp. mchwelet, f. 107,
etc; but odno ỳỳmchwoyl, f. 62<x (like MS. S); and d
ỳmchweilent, f. 63; a mchweilir, ff. 65, 676; d ruch-
weilant, f. 67; a mchweilassant, f. 90; a mchwelws,
f. 1006; a hnn oll a datỳmchweilies nnev, f. 1086;
a dchweiliassant, f. 141; pan ỳmchweilit, f. 129], and in
the prose textBuchedd yr anrhydeddusBedr,m Add. MS. 14,979
(16th cent.): ymchweylt a dwedd, f. 1586; onid ymchwelai,
f. 162; ymchwevlyd, f. 163; ymchweyldt, ib.; besides
ymchwelodd, f. 1576; etc. In the later MS. doydud and
dyweydyt occur (see 42; as to , cf. ib. onaddnt, f. 155;
dallt, gan e broclr, f. 1596; etc.


45. Tt results from the above quoted examples that ym-
lioelyd is a Middle- Welsh form surviving iu the South Western
dialects, whilst doedyd oecurs not earlier thau aboutthe lGth
century, and is not seldom met with in MSS. which for other
reasons sliould be attributed to the Powysian dialects. The
phonetic processes which are apparently common to both
must therefore be held to be different ones, and cannot be
further discussed here, though the phonetics of the modern
dialects, which are up to the present very incompletely
known, might perhaps illustrate them. Dyweidyd and
ymchweilyd probably contain the ei of the 3rd sing. dyweid,
ymchweil, which was analogically transferred into tliem at
the time when the 3rd sing. dywed, ymchwel, formed after
the model of cymnier, began to replace dyweid, ymchweil.
At that time the coexistence of dywed and dyweid caused
dyweidaf, like dywedaf, etc, to be formed. The modern
deidyd shows the group clw before vowels, simplified by the
tliiuination of w; we may remeniber dd from *dfod,
besides dwad (dyfod). The 3rd sing. Pret, of dywedyd is
dyfod, dywad, dywawd (Davies, Gramm.). On these forms
see Ehs, Rev. Gelt, vi, p. 17, who obseiwes that dywad, still
used inGwyuedd,is fornied " withthe modern preference for wa
over ico". Like gwared gwarawd gwares, an s-form of dywei 1
dywawd existed also (dywes-), on which see above, 32.

I. Tiie Veimj Si;bstaxtive.

4G. The forms <>f wyl', oeddwn, compounded with yd-,
ytt- 3 and with the preceding verbal particles ydd-, yr- (Xcnss.
-\). 551), show some phonetic and dialectal peculiarities not
mentioned in Zeuss. 5Td-, ytt- d some South-Welsh MSS.
from the L5th cent. downwards, and iu the later colloauial



language, generally appear as od-, ott-, sornetirnes wd-. 1 ( !f.

from Brut Gr. ab Arthur, in MS. Oleop., 135 (the only Middle-

Welsh MS. in which I found these forms): val r ottoedwn,

f. 60; thra ottodit, f. 73; gwr a ottoed, f. 90; r

ottoednt, f. 60&; a phan ottodnt, f. 10; marwo nychdod

ir odwyf, Eobin Ddu, 1460, quoted Add. MS. 14,944, f. 134.

In the Gwentian part (Huet) of Sal. N.T., very seldom: yrodoedo 1 ,
f. 386a. n forms occurring in the Gwentian Add. MS. 14,921
see my Beitr., 41 (odwi, ody, ni dody, ny dodi, nydody, odyn,
ny dodyn, odoedd, odys). In Marchawg Crwydrad (printed from
a South-Welsh MS.): lle nad ody, lle na bu a lle na bydd byth, p. 2;
er nad ody aur ac arian ond . . ., p. 148; y rhai a odynt yn eu
harfer, p. 145; ody duw yn 'wyllysu, p. 144; nid ody yn gwneuthur,
p. 143; yr arglwyddes Fenws yr honn ody yr holl garedig gariadon
yn foddlon iddi, p. 141; petheu .... drwy yr hwn y byd odys yn
ei drallodi. L. Morris, Add. MS. 14,023, f. 132//: y in South-
Welsh pronounced as o in Cdym, North-Welsh ydym, di, North-
Welsh ydyw.

In modern dialects:

Dimetian: a odi chi yn meddwl, Ser. Cymru, i, p. 429; a odi
Mari . . . ., p. 231; a odi chi yn dost? p. 251, etc. Gwentian:
Ehys, Lectures, 2 p. 45, gives oti, from a part of Gwent, that in
which the medise are pronounced in such a manner as to be
commonly written tenues, cf. jocal = diogel, in Neath, etc. Aber-
dare: odi (Y Gweitlwr). Llanelly, Cyf.yr Aelwyd: ydw, p. 68; ydy,
ydach chi, p. 20; besides ody? answered by odyn, p. 68. Y Tyw.
s'r G.: odi e'n ddrwg, ii, p. 66; oti chi, i, p. 154; otuchi, pp. 94, 117;
neu otu chi, p. 117; otuch I, p. 94 (ydu chi, b.); otus ddim . . . .,
p. 96. Monmouthshire: Y Bedyddiwr: otw, otw, viii, p. 174; nag
oti, p. 107; odi, p. 174; nag odyn, p. 174; etc.

Wd.: Seren Cymru: yr wdw i, iii, p. 264; wdw I, i, p. 232;
Gwron Cymreig: rwdwi, wdw i, 20, 5, 52; ib. wes, wedd, rwedd,
trwed for ces, etc. Carnarvonshire: mi rwdwi, Yr Arw., 9, 4, 57;
yr wdw i, ih.; rwdwi yu deud, 11, 12, 56; ac yr wdw ina yn
cael . . . ., etc. On similar phonetics in other wordsseemy Bcilr.,

1 yd before consonants occurs frequeutly in the old poems and in the
so-called Gogynfeirdd. Cf. i'rom Dr. Davies' MS. (Add. MS. 14,8i
e.g., yd vernir, f. 55a, yd rotir, y gelwir, yd gedwir, yt (= ydd) ergryner
f. 56a, mal yd glywir f. 706 (Cynddelw), etc.



40-48; cf.,e.r/.,bychan, bachan,bochan,bwchan (Wogan,Vwghan,
in the English part of Lhjfr Arhau, pp. 55, 56); to the forins of the
Old Welsh Catguocaun, given /. c.,40: Cydwgan, Ewdogan, Kodw-
gan, Ewdwgan, Eodogan; add. Eydogan, Ll. Achau, p. 22, Eadogan,
p. 15, Eadwgan, p. 15; etc.

47. Davies, Dict. and Gramm., p. 182, gives ydd-instead
of yr- as the Dimetian form. So also Pughe, in Coxe's
Monmouthshire: Gwent. yddoedd = Yenedot. yroedd; Dosp.
Edcijm, 824: yd- 3 ydd- are " more generally" used in South-
Welsh, y-, yr- in North-Welsh. This dialectal difference may
be seen in the t"\vo texts of Gruff. al> Cynaris Life used in
Myv. Arch.,\o\. ii; cf. ac yna ydd oeddynt, 2 p. 723 (South-
Welsh), yr- (North-Welsh); ydd oedd, 2 p. 724: yr- (Xorth-
Welsh); ac yny lle ydd annogasant wynteu, p. 724: yr anno-
gasant hwytheu; also pp. 730, 731: ydd erchis yr archodd;
etc. But yr- is not wanting in South-Welsh' texts and
modern dialects,at least as they are written in Scren Cymru,
Y Gicron Cymrcig, etc. These different forms are due to
the generalisation of mutations incurred before certain
consonants, like the different forms of the article in
Welsh and Breton. Yd-, notwithstanding its frequent use
in the form of od- in South-Welsh, is said by Davies,
Gramm., and others to be more generally used in North-
Welsh. Cf. Davies, Gramm.: Pass. ydys; poet.. Dimet.,
Powysian, ys; poet. also ydis. Hughes, 1822: South-
Welsh wyf rather than ydwyf. lowlands, Gramm., 4 p. 7G:
South-Welsh wyf, pl. ym, ych, ynt; North-Welsh ydwyf,
etc. Y Geninen, iii, 19, Glamorganshire: shwd i chi = sut
yr ydych chwi.

48. The prohlem of the origin of wyf and oeddwn inWelsh,
ouf and of, oann and en, etc, in Breton and Cornish, has
been very differently treated by Ilhŷs, Lccturcs, 2 p. 234, R. C,
vi, p. 4D, u. 1, and by Stokes, Kuhris Zcitsclu\, xxviii, p. 101
->"/. I am of the first opinionof the former, that these forms
belong to the root es-; ys = Ir. is; 3rd pl. iut, ynt = *enti 'or



*s-enti, causecl by *e- s- m . . . of the lst pl. From the very
early periocl when the vowel of the root, lost in the plural, was
reintroducecl in the lst and 2nd pers. pl. in which persons it
was afterwards altered, as the Cornish and Breton fornis
show the singular is supposed to have assumed thematic
flexion: *esemi, *esesi. Perhaps the assumption of another
analogical process, of the retransgression of the e of the plural
into the singular, or of the * (from the augment c + c of es-)
in the imperfect,would also meet the difficulties here existing:
*esmi gave *em or *ym (hence -m in buum bm, I was,etc);
*esi gave *ei, *esti gave ys, which has been kept; *e-esmi
(or *e-em) or *m, *si or *i, *sti or *s gave *wym (altered
later into wyf, like all other verbs), *wy (later wyt, with t of
the pronoun), oes, which still exists. An argument in favour
of the theory of thematic flexion is furnished by the 3rd sing.
yd-y, ycl-i, Corn. us-y, Bret. ecl-y, if -y is from *eset, and y w,
Corn. yu, eu, Bret. eu, is to be explained by the supposition
of an afxed pronominal element (Ehŷs). However, *set
would give *wy; and nyw-: nwy-, y'w (do), etc. caution us
against denying the possibility of *wy becoming yw under cer-
tainyetunknown conditions, ataveryremoteperiod,it is true:
cf. the Corn. and Bret. forms. These forms are analogically
transformed during their history in Welsh itself. J. D. Rhŷs,
Gramm., 1595, gives the pl. ym, yn (like all later plurals,
following the pron. ni), ych ancl ywch, ynt, Ywch, ydywch,
yttywch are ascribed byDavies to the Dimetian dialect and the
poets. Dosp. Ecl, 653, gives even lst plur. Dimet. wyn,
2nd. ywch, ywch. The modern South-Welsh dialects, how-
ever, use ych. Ydem, ydech, are the common North-AYelsh
forms, identical in termination with the modern Pres. Sec. in
-em, -ech, the only terminations of this tense used in these
dialects. So probably when -em was generalised in the Pres.
Sec. (see 17) ydym followed this model, like wrthachi for
wrthywch, etc. See Obs. on the Pron., 40.


Cf. Add. MS. 15,059 (18th cent., North-Welsh): pa beth ydachi 'n geisio, ai ceisio rychwi 'r fargen . . . ., f. 223b (ib.: nis gwn
i beth a ddywedach, pe gwelach; lst sing. a gefis, etc).

Venedotian, Sweet, p. 449: dw dwy dw, wyt wt, di di yw? (măy mă, ys, sydd sy); dan, dach, dyn. Yr Arw.: ydw, 18, 9, 56, mi rydw ina, 17nettlau-, 7, 5G; tydwi, 9, 4, 57 (I am not); yr wit ti,
11, 12, 56; y rw ti, 17, 7, 56; nid ydan ni, beth ydach chi, 18, 9,
56, etc. Powysian: pwy ydech cliwi, Y Cyf dyfyr, 1883. In Cab.
few. Tomos, a is written for final e in this verb, as in Carnarvon-
shire, though it is ordinarily written e in other words. Cf. ini
rydw i'n cofio, p. 7; nieddwl rydw i, p. 15; yr ydw i'n gweld, p.
18; ne dydw i'nabod; mi rwyt ti, p. 14; ydi 'r llyfr ddim yn
. . . ., p. 11; ydan ni, p. 77; 'r yda ni, p. 135; ond 'r ydach
chi, p. 135; mi rydach chi, pp. 19, 53; ai dydyn nhw (nad),
p. 33.

Cann. y Cymry, 1672: rwi 'n tifedd, p. 276; ar sawl wi 'n nabod
etto, p. 337; 2nd plur. ych.

Dimetian, Ser. C.: yr w'i yn meddwl, i, p. 212; w'ine, ii, p. 184;
beth w'ti yn, f. iii, p. 142; shwt w'ti, ii, p. 262; yr w'ti, iii, pp.
305,603; i nw, i, 251. Gwentian, Aberdare: dwi ddim . . . .,
ond w i yn meddwl, Y Gw.\ Llanelly, V Tyw. 'r G.: dw'i ddim, i,
p. 94; ond dych chi ddim, 1 chi, ib.\ a pha beth i chi 'n ydu
chi, i, p. 134; nad i chi, etc. Monmouthshire: y'ch chi, Y Bed.,

49. Tn the modern dialects, instead of oeddwn, etc.,

forms are used, given by Rowlands, Gframm., 2 248, as own, oit,

plur. oem, oech, oent. The earliest reliable examples of theiu

which I know of are 'roem (oeddem), Gann. y Gymry, 107-,

1.:!74, and South-Welsh oen, " they were" North-Welsh

oeddyn, L. Morris, Add. MS. 14,923, f. lUa; Williams Pant

y Celyn: o'wn (quoted in Y Tracth., 1870, p. 413); o'ent in

tlie magazine Trysorfa Gwybodaeth, etc, Carmarthen, 1770

(ina passage reprinted in Y Traeth., vol. iv). A oeint, O v wen,

Lttms, p. 304 (Crwcnt. Cod<\ MS. Y), and y ddoyent, /J. Acnw,

l>. 17 (ib., y doyeth = doeth, ]>. (iU), are of conrse doubtful.

Dimetian, Seren Oymru: own I, i, pp. 233, 272; yr own I, p.
232: ond'down I, U>.\ o'et ti i, p. 373; pan o ti iii. p. 184; yr o ti,

|i. 142; yr o ni. p. I 17; \v chi, p, 22; yn well nagonw r dyddie
basodd, yi- o'n awy i, p. 292 (er aad o <is i, p. ^7.'!). Seren (lomer,
:!'i: yr oet ti, pan oent, p. 362; o'i chi, p. 268.



Gwentian, Aberdare: yr own i, Y Gweithiwr, 1858, 5; oen nhw,
ib. (odd e, dos = nid oes, ib.). Llanelly: pan own i, paham o' it ti
'n chwerthin, Cyf. yr Ael., i, pp. 68, 32; Y Tyio. a\ 67.: yr own
I, oe' chi i, p. 94; oe' nhw, p. 96. Neath: own i, oit ti, dd, oin nw.
Monmouthshire: dyma 'r peth own i'n hela ato, YBed., 8, p. 106; o'wn
i, p. 174. Ebbw Vale: o'n i, lst sing.; o ni, o chi, P. Cymr., 28, 37.

Uorth Wales: in Yr Arw. I have noticed only oeddwn, written,
e.g., pen ouddan ni, 19, 11, 56; but Sweet, p. 449, gives 6n, lst sing.;
and so in Y Gen. Gymr. (Carnarvon): o'n i yn gwel'd, pan yr own i,
rown i, down i (nid), etc, 20, 5, 1885, p. 7 6 , etc.

These forms possess a strilring similarity to tlie Breton
oann, oas, oe, oa, oamp, oach, oant, and to the Cornish en, es,
o, 3rd plur. ens. Bnt their late appearance cautions us
against the assumption of their identity, and against attri-
buting the existence or non-existence of dd (from j) to
different accentuation, etc. I presumethat at thetimewhen
oedd was pronounced dd it was identified in termination with
the secondary presents in -odd, carodd, rhdd, fromrhoi, etc.;
and that o (as rho in rhdd), the supposed stem, was conju-
gated like another verb in the Pres. Sec.: o-wn, o-it, plur. oem,
oech, oent, like rhown, etc. -odd, it is true, belongs now to the
system of the s-preterite, and so we must probably attribute
to the inuence of oeddwn itself the fact that own, and not
*oais, was formed. Perhaps forms like dathoeddwn, 3rd
sing. dathoedd, pron. dathodd, also contributed to the forma-
tion of own, and not *oais. Oedd in these compounds, on
which see below, was further assimilated to the Pres. Sec.; cf .
y cawsei ef, marg. cathoddei, Sal., N. T., f. 148. Ym oeddy w
pro oeddwn, Add. MS. 14,944, f. 134.

50. As to the forms of the verb substantive commencing
with b (it is not safe to speak of them otherwise, since they
are attributed to so many different roots), the old forms of the
conjunctive and optative, imperative, present second, etc, are
being replaced in the modern language, at least as it is given
in grammars, by forms of byddaf, conjugated like any other
verb. Of this stem bydd- forms with o occur in the most



recent centuries and in the modern dialects, nearly all con-

junctives. Cf. bythoch or bothoch = byddoch, "that you

may be", Ihŷs, Lect., 2 p. 232. Th in the conjunctive is due to

the old accentuation, which also causes teccach, teg, etc. So:

hyt pann vethont, Ll. Gw. Bh., p. 289; pan vethont, MS. Tit.

D22, f. 156&; tra vythw vi, Add. MS. 14,986, f. 25; bythwyf,

rhothwyf, Davies, Gramm.

Cf. Add. MS. 14,898 (17th cent.): lle botho ych ffansi, f. 74;
Add. MS. 14,938: pen fotho, f. 6Sa; Add. MS. 31,060 (18th cent.):
fel y botho gwiw ych caru, f. 137a. Hope, 1765: botlrwi, p. vii;
fodde, p. 5 (byddei). Cahan f'ew, T.: tra botho nhw 'n dal. p.
132; and also pp. 14, 116, 117, 134, 135; am byd bothol ber-
ffaith, p. 278. Yr Amserau: nes botho bi, 4, 5, 48; gore pomwua
o blant fotho gin bobol i fund ono. 29, 3, 49; etc. Ono, dono and
ene, dene are equally often used in these ttythyrau , rhen fjfarmwr;
ib. neu bomtheg o byrsonied, 30, 12, 47, cf. i bwmthegarugen, Hen
Cymro in Y Gwron Cymreig, 6, 5, 1852 (Carmarthe). All these
texts belong to the North-Welsh dialects.

51. Davies, Gramm., gives the Opt. Eesp. Conj. as bwyf,

bŷch, bch, b, bom, bch, bnt. In Carnarvonsh. (Sweet,

p. 449): b, bt, b, bn, bch (?), bn. Boent and bwyntoccur

frequently in Middle-Welsh texts. Cf. MS. A, p. 77: elle

ebont = e boent; B, J, K: yr nep pyeufoent, p. 157; Tit. D2

(= B): ket etuoent, f. 22b; herwd e lle e boent endau

f. 32a; gwedỳ bunt, f. 37; Add. MS. 14,931 (=-#): a

gfarfoent, f. 17a; ŷn wlat d anhodnt ohone, p. 77,=

hanuoent, B, = henynt, K, Tj (Dimt Gode); lcynn bynt, p.

268, =: bont, <7, K; cyd bwynt in later texts. In modern texts

these forms, with tra preceding, lose the initial /: trech =

tra fech. The vowel of the proclitic tra was dropped like

ther pre-tonic vowels, and the group trf + vowel became

tr. Taswn for ped faswn (see below) is a similar case.

Cf. Cann. y Cymry, L672: trech (tra fyddtch), p. 130 (2); trech
liyw, jpp. 2 11, 295 (//>., lle bych dithc, marg. byddych); byth tro
ganto, marg. tra fo cf. In )' Traeth., iii, p. 8, tr'cch, tr'o are
said to be South-Welsh. From Williams Pant y Celyn tr'wi is
quotcd, Llyfryddiaeth, j>. 397; thisisfrom tra fwyf.



52. On the perfect (bm) see Bhŷs, Bcv. CelL, vi,p. 1 9 seq.
Both a and o occur in Micldle-Welsh in the terminations of the
lst and 3rd pl.; in the B. of Herg. and Ll. Gw. Bh. buam,
buant largely prevail. On the modern South-Welsh neo-for-
mations bues, buest, see 19. Another dialectal peculiarity
is buo (lst sing.) in the Dimetian dialects. Cf. Davies,
Gramm.: bum, poet. buym, buum; buo Demetarum vulgus;
also eutho, deutho, gwneutho (for euthum, etc). L. Morris,
Add. MS. 14,923,1 1336: South-Welsh buo, buofi, I have
been. Hughes, 1822: South-Welsh buo = bum. Spurrell,
Gramm., 3 189: South-Welsh buais, bues, buo.

Cf. March. Crwydr. (17th cent., South-Welsh MS.): mi gwybuo,
p. 150; pan gwybuo, p. 152; nid aetho, etc. Cann. y C: a wnaetho i
'n derbyn, p. 141; ar y wnaetho i maes o'th feddwl, o bwy nifer o
bechode wnaetho i 'n d' erbyn, p. 400; pob ffolineb ag a wnaetho,
i, p. 451; etc. Y Cymmr., iv (Iolo Morganwg): o fod gymmaint
ag y buof i maes yn y tywydd. Y Tyw. V G.: fe wnaetho I,
iii, p. 192; Monmouthsh., Y Bed.: y buo i 8, p. 106; a wnaetho i,
p. 108; etc.

Similar forms exist also in Carnarvonshire: lst sing. buom,
3rd sing. buo; cf. Sweet, p. 449: lst sing. byom, bŷm; 3rd
sing. byo, bŷ. Ehŷs says that bom i, which he thinlcs to be
old, is used in North- and South-Welsh (l. c. p. 20).

Cf. Yr Ariv. (Pwllheli): ac mi fuom yn hir iawn chwedun yn
misio .... 26, 2, 57; a gallwch feddwl wth a ddyudis i, y buo hi
yn fud garwinol arna i, ond ddyudis i mo hanar y ffrigwd wrtlio
chi, 17, 7, 56; mi fuo yr un o'r rhyini, 2, 10, 56; mi fuo fo, 13, 11,
56; ma nhw yn bobol nyisia (Eng. nice, superl.) a ffryindia fuo
rioud ar y ddyuar yma, 2, 10, 56; etc.

These forms (North-Welsh buom, South-Welsh buo =
buof, fnal/not being pronounced, and therefore not written
in these new forms, of which no historical orthographies
existed, and North-Welsh buo) are evidently formed after the
model of the conjunctive bf, pl. bom, boch, bont. Bum and
bu, isolated forms in the scheme of the verb, were either



directly made bu-of, bu-o, or o was at least combined with
bum, the North-AYelsh bu-o-m resulting.

A still more probable explanation seems to be the fol-
lowing: buom buoch buont lost their isolated sing. bm, bu,
and formed buof, buo, like caraf, cara; pl. caram, buam,cersom,
buom. In Neath, byo, byot, and byost, by a (ef); also etho i,
detho i, gnetho i, ceso i (cafael), 2nd sing. ethot and ethost,
dethot and dethost, etc, occur commonly. The doublets in
the 2nd sing. show the influence of the optative and also the
exact origin of attast, arnost, also wthtost, etc, quotcd in
Obs. on the Pron., and in common use at Neath.

53. The secondary preterite of the Verb Subst.,in Middle-
Welsh bu-asswn, bu-ysswn, becomes in modern Welsh, by
dropping the unstressed u, basswn, bysswn. Buswn also
occurs in theliterary language (bu-swn, like gwel-swn).

Of. Zeu8S, 2 p. 5C3, and Ll. Gw. Rh.: a uuyssyut ae nabuyssynt,
pp. 192, 206; Cleop., B5: pan uuessynt, etc.

Salesbury, N. T.: pe byesiti, f. lb\b. R. Davies: pe biase, f. 3046;
pe biasentwy, f. 340a Huet: by biasey, f. 379a; beBdes pe's
bysei, f. 946; ys gwybysei, besei, ff. 170, 220; ny besei, f. 1516;
a ddarvesei, f. 174; pe bysem, f. 38a; ny vesetn ni, f. 38; pe
besent, na wybesynt, f. 3606. Y Drych Christ., 1585: a fuassei,
p. 3; besides pei bysswn i, ni byssei, pei, ony byssei, f. 216; pei
bassai, f. 216; a phet fassci, f. 16; fe fassei dhyn, f. 216; etc.
Cann. y C.: ony bysse, ony basse, p. 139. E. Lhuyd, A. Br.,
[ntroduct.: bisse, pe bissent (South-Welsh).

Noih Welsh, Sweet, p. 449: baswn, basat, basa, basan. basach,
basan; y basswn i, Yr Arw., 30, 10. 56; piti garw na baset ti, mi
faswn i, Ca6. f'ew. T. South-Welsh: ni fyswn I, I byswn I, Ser.
C, i, p. 449; buse, na buse fe. a fusc chi, /b. Llanelly: fuse hi,
)' '/'. n'r G., , p. 65.
Moniiidutlish.: a fysse, os bysa hi, Y Bed.,
viii, p. 44, etc; y busai yn ddyinunol, etc.
(literary language).

54. After ped, " if ", the/ of fawn, fewn, fasswn, etc, is
dropped: petawn ="ped"- fawn. Even the vowel of the pro-
clitic "ped" is lost, and *p-t becomes /,the result being tawn,
taswn, <((., the common forms of the modern colloquial
language. In certain Middle-Welsh MSS. we find forms




which one might call " open forms", ancl which shall be
further discussed below; but the existence of forms like
pettei in the same MSS. shows that they are merely historical

Cf. B. ofBerg.: pei yt uen i yn dechreu vy ieuenctit, col. 745;
beyt uei, col. 638. Ll. Gw. Rh.: pettei a allei eu caffel, p. 226; a
phettut, p. 315. Y S. Gr.: a phei at vei o haearn pygyeit, pp. 195,
389; megys pei at uei milgi, p. 336; a phei at ueut, 2nd sing., pp.
392, 420; pei at uei marw, 64, 67; besides a phettwn inneu, p.
278; pettut, 27, p. 349; a phettei, 62, pp. 179, 243; a phettit,
p. 284, Pass.

Salesbury, N. T.: val petei, f. 1796; mal petyn, f. 63<r; megis
petyn, f. 212a; val petent, val petaent, etc.
Gr. Roberts, Gramm.:
petten, p. 14 (YDrych Christ.: pe y baem, D la; fal pe y baynt, C 16;
pe i bysswn i, 3). Add. MS. 14,986 (16th cent.): pe dfai e, be bai.
Add. MS. 14,921: fal bebai, f. 376; fal bebay, f. 276; fal be bai,
f. 19a; fal betai, f. 15/;; fel betai, f. 46; betaichwi, f. 446; a
ffetasse, f. 31 (fal bebassent, f. 45a). Cann.y C.: peteitin gwybod,
p. 505; pyt fae, pp. 260, 545.

In Grammars: pettwn, pettit, pettai; pettym, -ych, -ynt (Row-
lands); petaem, - aech, -aent (D. S. Evans).

Carnarvonshire, Sweet, p. 449: taswn, tasat, tasa, ta; tasan,
tasach, tch, tasan, tn. Yr Arw. \ taswn i, ta titha yn gwbod, wel
dachi yn . . . 17, 7, 56; dae ti yn myned, 15, 1, 57; wel, dasa . . .
20, 8, 57; etc. Merioneth, Caban feiv. T.: wel, tawn i, tase r
llyad wedi syrthio trw'r to, p. 61; etc. Carmarthenshirc, Ser. C.:
tawn I yn gwbod, 'tawn I, iii, p. 382; ta, i, p. 292.
dialect: fy licwn tau'chi 'n esbonio, Y T. a'r G. i, p. 1 17; ta chi, ii,
p. 67; etc.

55. As pei at-,owing to its non-accentuation,becomes ped
in the later lauguage, so pes (if) must be supposed to be a
proclitic contraction of pei as-, pei ys-, in which combination
s is an infixed prononn; cf. pei as gypn, mi ae dywedn,
B. of Herg., col. 835; Y S. Gr., 2, 26, 46; etc. In Gann. y
G, 1672, pes is still printed pe's, p. 414. Davies, Dict.: pes
from pe ys. Salesbury, N. T.: pe's bysei hwn yn pr., ys
gwybysei ef; pe's adwaenesoch vi, ys adwaenesoch vy tat
hefyt, f. 145&; dywedaf ychwy pe's tawei i'r ei hyn, ys llefai




yr main, marg. ceric,the South-Welsh word (pe tawai y llefai,

ecl, 1873); etc. Compare also neur, neucl, neus ancl nicl, nis.

Though, with perhaps one exception, no "open" forms are

retained, ot and os, or, if, besides o (aspirating), if, certainly

contain the same elements as the second parts of pet, pes.

Or, probably from *o yr, is given by Davies and others as

South-Welsh. It is not so easy to separate the spheres of

ot, os, or, as those of pet, pes. Cf. B. of Hcrg.: namyn ti a

gyr dy ty a eill dial owein or llas, neu y ryclhau ot ydiw

ygkarchar. ac os bu y dyn gyt a thi, col. 642. As I pointed

out in my Bcitr., 55, o (aspirating) lost in certain positions

its final tenuis, like a and ac (and), but *oc seems still to

exist in agatfydd, agatoedd, an olcl compound which has more

or less lost its verbal character, and become an adverbial

formula. Otherwise, *o atfydd, perhaps *otfydd, would have

sprung up. That ag at- is really the representative of a later

o at-, is perhaps supported by the once-occurring o gattoeth

(MS. S. f. 63a: Pob gad hagen gann dyg k6byl a vyd digan

[MS. digiga6n] yr g6ad6r ac yr reithyr o gattoeth kynny bo

g6ir), by the occurrence of a, at, as, ar (if), common, c.g., in

Salesbury's N. T., etc, and perhaps by osgatfydd, ysgatfydcl

(Spurrell), forms in which os (if), unstressed ys, was reintro-

duced, at a time when agatfydd was no longer known to

contain a word denoting if

Cf. acattoed, ac attuyd, B. of fferg., col. 673; agattoed, Ll. G>r.
Ih.,-p. 85; ac accattoed, p. 2G3; agatfydd, scatfydd, yscatfydd, J.
D. lhys, Gramm., p. 106. Atfydd alone occurs, e.g., in B. of I/( rg. t
col. 740: a phei mynnvt gyuoeth eiryoed atuyd y kaffut ti hnn.
)'.s7. Gwl. leuan \\m/., p. 331: rac attoeth cayu y ddayar arnaw;
atoedd (otherwise), Sp., Dct., etc. Adfydd is used quite as an
adrerb, and even compounded with di-; cf. diedfydd, ccrte, procul
dubio, Davies, Dict.


56. The verb substantive exists in all its forms in com-
position witli the prepositions gor-, can- (ar-gan-), d-ar- and



" han-" (on whicb see my Obs. on the Pron., 82); with the
unexplained pi- or pieu-; and with tlie verhal forrns (not
stems, of course), gwydd-, gnaw-; as well as in the Part.
Pret. Pass. with forrns externally similar to gwydd- in all
verbs (e ducpuet, MS. A, p. 48, etc). Other apparent com-
pounds, like taluaf, MS. A, p. 58, o deweduuant, p. 69,
dywetpynt, MS. F, p. 390 (dywettynt, MSS. G, U), are
extremely rarely met with, and if they are not errors, they
are occasional results of the compounds with bwyt (buwyt),
-pwyt. The early compounds are in later times con-
jugated like byddaf, a regular verb, or else the first elements
of the compounds, together with the initial b, modifed
according to the consonant preceding, are erroneously
abstracted as the stem of a verb, otherwise conjugated
regularly. So gwypwyf is gwyddbwyf; but gwypwn, Imp.
gwyped, are formed from the pretended stem *gwyp-. So
from adnabod: adnappwyf, adneppych, adnappo, etc.; but also
adnappwn, adnapper. E-f becomes r-ff. So arose darffwyf,
gorffwyf, which were imitated in hanffwyf (and hanpwyf).
Hence hanffwn, hanffer, hanffasid, and even hanffaf, henffi, 3rd
sing. henffydd (*hanff is likely to have been avoided), pl.
hanffwn, henffwch, hanffant (Davies, Gramm.). Cf. yr nep
y gorffer arnaw, Ll. Gw. Rh., pp. 34, 43 (y goruydir arnei, p.
34; a orffo, p. 34); pan ar ganfer gyntaf, Add. MS. 14,912,
f. 17b. The kind of infection of the initial consonants in
the second element of these compounds varies, as shown by
the different degrees seen in can focl, gwy bod, gor ffwyf,
which were imitated by the other verbs, c.g., hanfu and
hanbu (like gwybu), canffydder, canffer, and canfydder, etc.
57. Yd henwyf o honei, yd henyt tithen, B. of Hcrg.,
col. 711. Cenyw, D. ab Gw., Poems, p. 205. Amdenyw is
evidently an imitation of deryw, etc.; cf. Dosparth Edeyrn,
939. Goryw is given, ib., 813. Ny dery, B. of Herg.,
col. 740; pa dery itti, col. 751. The common form used in

H 2



the B. of Henj., liowever, is derw, for examples ot' wliich see
my Beitr., 110 (cols. 568, 639, 642, 660, 661, etc.). 1 Fut.,
a derbit y euo, Ll. Gw. Bh., p. 268; Imp., derffit, B. of Herg.,
col. 566. The Perf. darfu becomes in the modern lnguage
daru, which is used with an infinitive following, especially
in the Northern dialects, to circumscribe the preterite of such
verbs as make little use of their s-aorists. Hnghes, 1822,
says: " In North-Welsh they are fond of darfu; beth ddarfu
iddo wneuthur?'' Cf. also Sweet, p. 445, and the dialectal
texts, in which be haru, ba haru is sometimes written for
beth ddarfu: be haru chi hyiddiw, Yr Arw., 17, 7, 56; ba
haru mi yn wir, ib.; be haru ti, Cab. few. T., p. 337. Cf.
myrolaeth, Stowe MS. 672, f. 230, etc.

58. Pi-, pieu-, in composition with bod, denoting to " pos-
sess", is one of the most enigmatic Welsh ẁords to me.
A few examples are:

MS. Calig. A13 (=C): ỳr nep pỳeiffo e da, f. 177?;; pye6, f. 182*.
MS. L: bieu, p. 240; biỳnt in M. B. of Herg.: p6y bieynt h6y,
col. 752; bioed, cols. 655, 662, 663, 668, 691, 775; bieuoed, col.
658; bieiuu, col. 775. Ll. Giv. Eh.: ni ae pieuvydwn hwy, p. 64;
bieiuyd, p. 81. Y S. Gr.: ny wydit pwy byoed, p. 337; pioed
gwr ef (whose man he was), p. 215; piwyt gwr di? (whose man
art thou?), p. 222; (y bwy yd oed gwr ef , p. 252). Salesbury, N. T.:
pwy bieivydd, f. 106>; ny phyeivydd ef hwnw, f. 231a; duw ei
pieu, f. 249n; y pieffont tymasoedd y bud, f. o85b (Huet). It
is still used, since I have read, e.g.t, pia, Cbfew. T., p. 7; waith
taw chi pia hi, Y Bed., viii, p. 107 (Monmouthshire); etc.

As to the etymology of this word, I can only say, that its
use in sentences like those quoted from Y Scint Greal is

1 In " A Treatise on the Chief Peculiarities that distinguish the
Cymraeg, as spoken by the inhabitants of Gwent and Morganwg
respectively", by " Pererindodwr",in the Cambrian Journal, 1855-7, derw
occurs in the following sentence: ydych chi wedi derw liau gwinith?
Odyn, ni gwplaon ddo (Monmouthshire and East Glamorganshire):
wedi darfod hau gwenith, niddarfyddson ddo (West Glamorganshire),
and: odich chi wedi penu (dybenu, darfod in Cardiganshire), hoi
gwenith? Odyn, ni benson ddoe (Dimetian dialects).



suggestive of pi- being a case of tlie stem of tlie interroga-
tive pronouns, and its combination with -eu- being similar
to that of ei in eiddo, if this be really froni *ei-iddo (see
Obs. on the Pron., 51, 52). The sense would be: cui est
and cui suum est (pi oed, pi eu oed). But I cannot
explain the phonetics of pi, also written py. *po-i gives
pwy, pw, and py; but i in pi-, pieu- is nearly constant in
the MSS. Cf. of course Cornish pyw, pew (pewo, a bewe, ty
a vew); Breton biou, byou, e.g., rac me biou, Sainte Barbe,
37. I have not yet seen Ernault's opinion on pieu in his
Middle-Breton Glossary.

59. On the perfect gwn, gwddost, gwyr, etc, of gwybod,
see Wny$,Rev. Celt., vi, p. 20-21. Sweet, p. 450, mentions the
2nd sing. gwst from Carnarvonshire, directly forined from
gwn. Old-Welsh amgnaubot, Oxon. 1: in the South-Welsh
dialects gnabod is used for adnabod; cf. Powel (Annotat. to Tit.
D22, f. 7): Dimetian nabod. In ISTeath: gnapod, 'n napod,
nabyddas, -ast, -ws, -son, -soch, -son; fe nabyddwd. Add.
MS. 14,986 (16th cent.): yny gnebydd di dy hvn, f. r o%;
Cann. y C: 'n nabod. Aberdare: yn cadw cnabyddiaeth a
yn gilydd, Y Gweithiwr, 1858, 5; yn nabod, yn gnabyddus
iawn, ib. (cn probably as in " cnawd, vulgo perperam pro
gnawd, Consuetum", Davies, Dict. I cannot discern whether
by a phonetic change or by a neo-formation, gnabod being
believed to be in the status nfectus); Monmouthshire: a
nabyddswn yn dda, dwy 'n nabod, Y Becl., viii, p. 106. Also
dynabod is written; Add. MS. 14,986: reol i dynabod, f. 16
(ydnabocl, f. 17; hydnabyddwch, f. 19; kyndnabyddwch,
f. 23a). Add. MS. 14,913: yddnabod (cf. hiil, ff. 'a, 4J, ib.),
f. 16; ay ddynabod y hynan ac y ddynabod y aiddio y hyn
rac aiddio arall, f. 16. Y Drych hrist.: dynabod, Cl
(ednabod, f. 17). Lewis Morris, in Add. MS. 14,944, f. 206:
dynabod in Anglesey = adnabod. Perhaps it is only a
phonetic variant of adnabod, ydnabod,like ymsangu,mysangu?



60. Adwaen, in the B. of Carm. once adwoen, as the

rhyme shows (see Ecv. Oelt, vi, p. 22), is from *at-gwo-gn-

*u-ve-gn-, corresponding to the Irish aithgn, from -ge-gn-.

In the B. of Herg. and Ll. Gw. Eh. adwen is the conmton

form of the lst and 3rd sing.

Cf. B. of H> >■;/.: atwen, Shene, pp. 261, 268; col. 772, 781, 8376.
Ll. Gw. Eh.: nii ath atwen, pp. 73, 89; yd atwen i, p. 97; mi a
attwen, p. 101; 3rd sing. ani hattwen, p. 235 (besides yd atwein i,
p. 102; hyd yd atweun [cf. deutb, or Icg. -v/enn?] i, p. 95; nys
ctwenwch, p. 235; canyt etweynwch chwi, p. 33; attwaenat, p.
10G; ny ani ydwaynat ef, p. 205; yny ettweinit, p. 217).

Probahly the old ord sing. edwyn, frorn *ed wwyn, the

result of ad wo en + a slender vowel, *wwy becoming wy,

was the source of atwen, forrned like gwel-: ord sing. gwyl.

Wy in edwyn was v+y; *edwn never occurs.

K. The Eoot *ag in the Vebb, and its Compounds.

61. Ag- simple, and ag compounded with do- and " gwn-",

furnish part of the forrns denoting to go, to corne, and to

make. As to gwn- 1 have not found mentioned in the proper

place in Zeuss 2 (though occasionally quoted,p. 600) the older

forms gwen- and gwan-, which are of very frequent occurrence

in the oldest Venedotian MS. A., but which are extreniely

seldom met with in other texts.

Cf. in MS. A: gueneuthur, pp. 6, 10, 11, 12, etc. (24 times)
gueneutur, pp. 14, 18, etc. (six times); gueneythur, pp. 15, 892
gueneythr, p. 18; regueneutur, p. 43; besides guneuthur, p. 53
guneyhtur, p. 66; guneythur, p. 67; guneithur, p. 62; gunehui
p. 65; gneuthur, pp. GG, 388; gnethur, p. 69; gneutur, p. 393,
gneihiur, p. 65; gneihur, [>. 65; gneisur, p. 62 (in certain parts of
tln.s .Ms. /,, ///, s are tmtten for th; s occurs also in the fragment
in older orthography in MS. E; kefreisial = cyfreithiol, p. 407);
ineuthuredyc, p. 389; 8rd Bing.: guenel, pp. 6, 10, 13, etc.
(22 mes); besides gunel, p. U; gnel, pp. 20, 22, 155; 3rd sing.,
guana, pp. 10, II, 57; guna, p. 1U; a guanant, p. 46; Imp. guanaet
l'-;;:I1 i ''' lenelheŷ, p. 1; guenelỳnt, pp. 153, 389; besides

gnelei, p. 63; guenelhont, p. 161 j guanaeth, p. 55; guanaet, p.
154; gnaeth, p. 55 j gnaet, p. .'. I; gnayth, p. 66; gueneler, p. 10,
31, etc. isix times) j geneyr, pp. 3, 37.



Add. MS. 14,912, Middle-Welsh, containing Meddygon Myddfai,
etc.: awyneler, f. 296; awyna, pop petli a wynel, a wyneler, a
wỳneler, pop gweith awynelych, f. 30a; a wyneler (twice), f. 30;
a wynelych, a wnelych, f. 31a. These are the only instances
occurring in this MS., which, by the way, abounds in South-Welsh
dialectal peculiarities; cf. gedu hỳnuỳ, f. 40; gedu darfo, f.
22b; o honu, f. 27b (3rd sing. fem.); besides neb auenno, f. 27b;
echedic, f. 40&; efuet, f. 40; a oba, f. 46a; dechrer, f. 31Z>; pl.
erill, ff. 36a, 37a; dffr, f. 40; dwfr, f. S7b; llester, f. Ub y
pym mylyned, f. 30a; degmylned, f. 31a; auu, f. 406; yn hoeth,
ff. 4lcr, 41t; llyssewyn, f. 36; giewyn, f. 36; bola, f. 39a; mis
whefrar, f. 196; hwer, f. 16; y wechet dyd, f. 30a; cawat,
f . 28a; anaddyl, f. 94a; hiddyl, ff. 437>, 48a; dan waddneu, ff. 26,
34; breudydon, f. 31b, breidydon, f. 30&, breuddon, f. 30,
breithyt, f. 31a, breuthton, f. 30J; dwesbwyt, f. 34&, sp,
f. 41a; regedac, ff. 145, 18a; uỳnd, f. 23a; etc.

In other texts I have only found that in MS. 21, p. 516, is
printed: my a dystyaf na wanevthost nev my dystyaf a dywedaist
(corruptly in MS. M, na wanay tyst).

Salesbury, N. T, f. 297a: eithyr ys da y gwanaethoch ar ywch
gyfrannu i'm gorthrymder i, marg. am blinder (= er hynny, da y
gwnaethoch gydgyfrannu 'm gorthrymder i, ed. 1873); gweny-
thei Marc 15, 8, is a printer's error for gwneythei. Athrav. Grist.,
p. 18: pa beth a uennaeth Crist yn arglw. N. pan dyscyno^ ef i
uphern. Add. MS. 19,874 (17th cent.): Gwrandewch bawb a
dowch ynghyd i gyd mewn ysbryd ddiddig, etc, f. 112a: Ir Tad
ar mab ar ysbryd glan gwanawn ine gan o volian val y dylem
bawb dan go roi iddofo r Gogoniaut, ib.; ysbrvd, plur. bigelvdd,
a ddoede, doeda, vddvn, golini, etc.; gwanawn was pronounced as
a monosyllable (gnawn or gnwawn).

Rees, C. Br. Saints, p. 225: gwynei, gwyney. These do not exist
in the MS. (Tit. D22), nor does wam. ntiyr, ib., p. 105, note, from a
Jesus College MS. (Lyfr yr Angcyr, No. 119, where y bnneuthur
is written in that passage).

62. Gwan-, gwen-: gwn- exhibit the sanie loss of the
pre-tonic vowel as gwar-, gwer-: gwr- ancl gor, from gwor-
in gwarandaw, gwerendeu, gwrandaw, gworescyn, gorescyn;
and indeed gwan- is nothing else than gwar-, gwor- (ver-);
as the other Brythonic languages and the Welsh goreu, goruc
show. iV has not yet been explained (see Rev. Celt., vi, p.
31); I think that it first arose in compounds w T ith el-, forms
like *gwer eler, etc, being altered to gweneler to avoid the



awkward recurrence of r and l. Though the orthographies
gneuthur, etc, in MS. A, show that gwn- was then pro-
nounced as in modern times, the orthographies guen- and
guan- in this MS. and in the otlier Middle-Welsh MS. are,
in my opinion, historical ones. Not so the few later ones,
which may be so-called "inverted orthographies", imitations
of gwar-: gwr-, both being pronounced gn-, gr-, or gnw-,
grw-, as Sweet describes these groups of consonants in the
Carnarvonshire dialect.

63. Doeth, daeth, deuth are all frequent in Middle-Welsh
MSS. Doeth is said in K. Jones' Worhs of Iolo Goch, 1877,
p. 13, n., to be S.-W. 1 Daeth, deuth are imitations of aeth,
euthom (like caer, ceurydd, maes, etc). MS. A: e doeht, p.
50; e doythant, e doetant: ethaethant, p. 50. Only by full
collections from more extensive texts could the chronological
and dialectal differences in the use of doeth daeth deutli
be shown.

Doeth, the form most used in the best Middle-Welsh MSS., dis-
appears iu later texts. From Sal., N. T., cf. y ddaethant, f. 214/;;
na ddauthym i, f. 186; pan daethesei, f. 2016 (wnaethesent, f. 233a);
athant, f. 199a; dathant, f. 200. Huet: deyth, f. 3896; deythont,
f . 388a (eithont, f. 393a; mi eythym, f. 3846); daythont, f. 3816;
y ddoyth, f. 382a; y ddoethont, etc.

64, In the compounds of aeth-, doeth-, gwnaeth- with
forms of bod (aethoedd, etc.),in the B. of Hery., Ll. Giu. Jlt.
and the later Middle-Welsh MSS., ath-, doth-, dath-, and
gwnath- are very common.

Cf. B. of Herg.: athoed, cols. 715, 78G; dothoed, cols. 662, 706;
dathoed, cols. 658, 659, 673, 705, 786; dathoedynt, col. 841; a
wnathoed, col. 723; nathoed, col. 839; gnathoed, col. 802; a
nathoedit, col. 732. Ll. Gw. Rh.: athoed, p. 98; athoedynt,
|,|.. 108, L16; dothoed, pp. 160, 201; dathoed, p. 116; dathoeth,

1 L. Monis, in Add. MS. 14,731 (Poems of Lewis Glyn Cothi),
. I !■/. ootea: doetfa prodaeth,;i Bolecism (c).



p. 55; ry wathoed, pp. 103, 263. Add. MS. 19,709: y dothoedynt,
f. 486; dothoed, ff. 286, la, 66a; rydothoed, etc. Y S. Gr.:
gwnathoed, p. 219; gwnathoedwn i, p. 198; na wnathoedut, p.
274, etc. Cleop. B5: r pan dathoet paganieit, f. 566; ỳno
dothoed G., etc. Darcs Phrygius: a athoedynt, f. 225?; 2; na
dothoed,f. 236al, 240 1 (doethoedynt, f . 226a 1; doethoed, ib., f. 2236
1, etc). Pughe (in Coxe's Monmouthshire) gives dothoeddynt as
the Gwentian, daethynt as the Venedotian form; this relates to
the Middle-Welsh Gwentian MSS. In the two MSS. of Buci.
Gruff. ab Cynan (Myv. Arch.) cf. rhyddothoed, 2 pp. 723, 731
(North-Welsh MS.: y daethai); a ddathoeddynt (North-Welsh; a
ddaethant), p. 725; ac oddyna i doethoedynt (North-We]sh:
oddiyno ydaethant), ib.; rhyddoethost (North-Welsh: y daethost),
p. 726; doethant (North-Welsh: daethant), p. 724 (twice), etc.

65. Davies, Gramm. (cf. Zeuss, 2 pp. 590, 591), gives
etliwyf D. G.; eclclwyd D. G.; ethyw, vet. poet., eddy w; deddy w
D. G.; dothyw, Bleddiu Fardd, 1246; doddyw, y dydd eddyw
in Powys = y dydcl a aeth, qui praeteriit; gwneddwyf D.
G.; gwneddyw D. G.; gwnaddoedd D. G.

Cf. from Add. MS. 14,869 (John Davies' transcript of the
Gogynfeirdd): Meilyr: ny dotynt, f. 185a. Gwalchmai: ny daw
ny dotyw, f. 1886; pan dotwyd, f. 197a; ethyw, f. 1886; na
dotyw, f. 136. Cynddelw: nyd athwyf, f. 43a; nyd adwyd, nyd
etiw, f. 43a; athwyd, dotyw, f. 58a; dothwyfy, f. 61a; dotwyf,
dothyw, f. 62a; ny dotwyf, f. 796; neum dotyw, f. 7la; neud
adwyf, f. 856; etyw, ff. 74a, 756; nyd etiw, f. 83a. Gwynnuart
Brycheinyawc: ethwyf, ethynt, f. 1146. Einyawn wann: detyw
f. 95a; neut eddwyf, f. 99. Bletyntuart: nyd etiw, f. 2086.
Llywelyn Fardd: Ysgwn cwdedyw ny un cwdaf, dothwyf. Pryd.
Bychan o Ddeh.: athwyf, f . 143a; ethym (twice), ib. Pryd. y
Moch: dydotyw y dyt, dydel, f. 1606; cf. dydaw, dydeuant, dybyt,
dybu, dybwyf, dybo, Llyw. Fardd, f. 1186; dotyw, f. 194a, etc.
T= dd in this MS., which is copied from a much older one; cf. f.
2376: finis. 16 April. 1617. Totum scripsi ego Jo. Davies. Hyd
hyn allan o hn lyfr ar femrwn, a scrifennasid peth o honaw
ynghylch amser Ed. 2. ac Ed. 3. fel y mae 'n gyffelyb: A pheth
ynghylch ainser Hen. 4. [a. 6. added later\. Yr hen lyfr hwnw
a fuasai yn eiddo Gruff. Dwnn, ac yn eiddo Huw llŷn, ac yn eiddo
Rys Cain ac yr awrhou sy eiddo Robert vychan or wen graig ger
llaw Dolgelleu. Scrifenyddiaeth y llyfr hwnnw oedd fal hyn yn y
law hynaf.



Cf. also B. ofHerg.: kyllell a edy ym byt, col. 812; ys ethyw
gennyf deuparth vy oet, col. 813; yny del gereint . . . . or neges yd
edy idi, col. 780; neut athoed hi heiba, col. 715, = neur ry adoed
hi heibya, col. 716; ydodyf, col. 718; y dodwyfi, col. 718;
dodyf, col. 77o; ny dody vyth dracheuyn, col. 806; neur dethynt
ythmis, col. 411.
Ll. Gw. Ilh.: y dyd a ediw ar nos a deuth,
p. 166; y dyd a ediw ar nos a dyuu, p. 171.
A ethy, ny dodynt
(dodyw, B. of Herg.), Rev. Celt., viii, pp. 23, 25. In Dafydd ab
Gwilym's Poems (ed. 1789): deddyw, p. 4; gwneddyw, p. 429,
a wneddwyf, p. 115.

Though ae and oe were pronounced ā,
ō, in the South-
Welsh dialects as early as in Middle-Welsh (see my Beitr.,
79), it is not probable that this pronunciation should have
been so often expressed in athoed, dothoed, whilst in the
same MSS. hardly any other instances of
ō = oe occur.
The decisive point in this question is the existence of ethyw,
eddyw, which are formed from athwyf, like henyw, deryw:
hanwyf, darwyf. Hence eth-, edd-, dedd-, gwnedd- were
analogically transferred to other forms. The origin of ath-
and doth-, add- and dodd-, is obscure to me, if they are not
mere alterations of aeth-, doeth-, a and o being introduced
from f, doaf = deuaf, etc.

66. Deuaf has been explained by Ehŷs from *do-(a)gaf.
It has not yet been shown how the 3rd sing. daw and the
Ist sing. doaf and daẅaf accord with this assumption, unless
we inay assume a double treatment of g in certain positions
between vowels similar to that of *v in ieuanc: *iewanc,
iefanc, ifanc; of *w in taraw, gwrandaw: 3rd sing. tereu,
gwerendeu; and also in ceneu: cenawon; llysewyn: llysieuyn;
giau: ui<'\\ yn; etc. 1 Then it would become necessary to regard
deuaf and daw as old forms, and doaf, doi or deui as analo-

1 Eiddew, ivy: ar erinllya ac eiddo ỳ ddaŷar, Add. I\IS. 1 4,912, f . 236;
eido y ddayar, f. 58; eidyo y ddayar, f. 63a; eiddyaw y ddayar,
I t,918, f. 626: itsha, used at Neath, BUggest even a triple treatment of
the roup; cf. eisio: eisieu (lilceitshafrom ei</ieu d unesplained)

on oue side, and gieu: giewyn (ciddcw) on the other. So daw-: do-:
deu-? f course Cornish and Breton must also be taken into account.


"ical formations. As the reasons of this double treatment of

the consonants have not yet been made out, and as it may
depend on the quality of the following vowel or even on the
accent, one cannot trace the exact worldng of the analogy
arising from deu- and daw-, do-. Both occur in the oldest
Middle-Welsh MSS.

Cf. B. of Carm.: a doant, 18; doit, 35. B. of Tal.: o dof, 8.
MS. A: doet, p. 50; doent, p. 73. MS. L: doant, p. 228 (deuant
in J and Q); doet, p. 210 (deuet in J, P); ony deuant, p. 242
(doant in /, O, P). Harl. MS. 958 (= T): or doant, f. 24; ony
deuant, p. 33a. B. ofHerg.: doch, Imp., col. 782; doei, col. 665;
2nd sing. deuy, cols. 661, 774, 788; Iinp. deuet, col. 659; etc.
Ll. Gw. PJi.: doy, 2nd sing., p. 138; dowch, Imp., p. 152; na doent,
p. 170; doei, p. 226; and doy (3rd sing.), p. 207; ny doynt, p.
194, etc.; besides deuant, p. 226; deuit (Imp.), p. 51 (deuity gof it
dywedut yn let it come to thy remembrance to inform us. The
imperativein -it ceridduwfi, Davies, see 3, and cf. bid, bandid
occurs very seldom in Middle-Welsh prose; another example is:
y neb a gredawd mywn kelwydeu. madeuit duw udunt, p. 85).
MS. Tt. D22: doant, ff. 1, a; doe, f. oa (dech, f. 13b).

Only in MS. Clcop. B5, I have found on deuwei, f. 75a;
ndeuwei, ih.; a deuwant, f. 61. In the same MS. beuwd,
f. oa (bywyd), ancl gwas ieuwanc, f. 60, are written. Pro-
bably these forms were pronounced dewant, dewei. In Add.
MS. 14,912: ac 6nt a ddoant oll allan, f. 60&; ac 6y a
ddant (o later inserted), ib.; 6ynt a ddoant, f. 626.

67. From the 16th cent. dawaf is frequent in South-
Welsh dialects. Davies, Gramm., has deuaf, deui, daw (old
diddaw), deuwn, -wch, -ant; vulgo dof, doi, daw, down, dowch,
dont; Dimetian dawaf, clewi, plur. dawn, dewch, dawant;
Imp. deuedj vulgo doed, Dim. dawed; Pr. Sec. deuwn, etc,
vulgo down, doit, dait, doi, dai; pl. doeni, daem, -ech, -ent,
Dimetian dawn, dewit, dawai (da6ai, Myv. Arch., x ii, p. 84);
pl. daweni -ech, -ent; Pass. deued, doed, Dimet. daed;
deuir, doir, dewir. Dait, dai, daem, etc, are imitations of
ait, ai, like daeth and doeth. -i in the 3rd sing. of the Pres.



Sec. is indeed uiore frequent in this verb than in others (see
L3), but cf. doei, B. of Herg., col. 665; Ll. Gw. Rh., p. 206,
etc. Gambold (1727) has also deir like eir.

i f. also L. Morris, Adcl. MS. 14,923, f. 133: South-Welsh
deuwch, dwch = North-Welsh dowch. YTradh, iii: South-
Welsh dawaf, dewch, dawant; dawer, dewir = North-Welsh
deuer, doir; Infin. dawed; heuce dawediad (e.g., Seren Gomer,
i, p. 160) = dyfodiad.

Salesbury, N. T.: dawaf, f. 247/); dawant, ff. 726, 3346; Imp.
dawet, yr hon a ddoy, marg. ddawei, yn dawot, ff. 46, 66, 126,
'..!/*; dewot, f. 886; besides ua ddauy allan oddyuow, f. 76; y
dauei, f. 1156; a ddauent y mewn, f. 37a; pan ddeuawdd, marg.
ddaeth, f. 3596 (R. Davies), is seldoin met with. Y Drych hrist.:
na dhawei, f. 29; sy n dwad iti, B16. Add. MS. 14.921 (Gwen-
tian dialect): ny ddawant, f. 386; y dawant, f. 53; y dawan,
f . 57a; Inf. dofod, ff. 12, 416, and dyfod. Add. MS. 14,98G: a
ddawa, f. 37; dowch, f . 386; dywod, f . 346 (also given Ln Davies,
Gramm., as occurring " aliquaudo"). In the Poems of Iorwerth
I ynglwyd, in Y Cymmr., vii: I dawaf J, MS. 11, p. 194; y doya vi,
MS. L (Stowe MS. 672), a curious f orm; ib.: doyau, 3rd pl.,
f. 317 (misbound); Gwentiau dialect, cf. ny thjoedd, y dyoedd,
aissiay, aissoes, cailog, etc.; ddwad, p. 193 (see Obs. on the
Pron., 68, 69); and doro, p. 192 (dyro). Cann. y C, 1672: dawe,
p. 25; dewe, p. 30; doe, p. 27; o'r doi, marg. os bydd itti ddyfod;
Inf. dwad, p. 61. E. Lhuyd, A.Br., uses dad (= dwad). Williams
l'ant y Celyn, according to Y Traeth., 1870, p. 412, wrote dewa,
3rd sing., a form of dawaf, like the modern cara, doda, -a beiug
borrowcd from the verbs in *-agaf .

68. In tlie modern dialects df is usediu North-Welsh
and dawaf in South-Welsh 1; the infinitives (besides dawed)
are dd and dwad, dŵad. Cf. Yr Arw., 17, 7, 57 i clwad; also
in Ob. few. T., p. 7; etc. Both dwad and dd are due to
accentuation of the second syllable, y being dropped. From

1 Cf. Cambrian .Iminial, 1856, p. 216: Dim. o ble doest ti, o ble
deithoch chi: Glam. and Monm. o ble dest ti, dethoch chi; p. 251:
Glam. (ast), Mimin.: des: Dim. dathym? Detho i is used at Neath;
ib.: dŵa i (bo, (.//., na ddŵa ddim, )' Fellten 16, 2, 1870), dewi di (doip
di, Poutypridd), daw; Inf. dud and dŵad.



*dfod arose dod, froni *dwod (cf. Davies, Gramm., dyfod, ali-
quando dywod): dwad, dŵad, like cwad from cyfod (Add. MS.
15,038, f. 606: kwad i vyny; also in Add. MSS. 14,973 in the
same text, and 15,059, f. 223a; Y Cymmr., vii, p. 186: kyfod,
cyfod, kywod MS. S5, cywod MS. B) besides codi from
*cfodi. Wa for older wo, like dwad, dywad for dywod,
dywawd (he said); marwar, B. of Tal., p. 119, marwor, Mcdd.
Mijcldf., p. 91: marwar and anwar, dr, llachar, rhyming,
Gwaiti Ll. Glyn Cothi, p. 61; marwar, Davies, Dict.; etc.

69. The imperative dos is proved by the Corn. dus, des,
Breton deux, to be an old form, but it has not been ex-
plained. Does also occurs, perhaps caused by the 3rd sing.
doec. J. D. Rhŷs, Gramm., conjugates dos: ds di, dosset ef,
dosswn, etc, on which forms Davies remarks that they are
" sine ulla auctoritate ne vulgi quidem". Since moes is
ecjually obscure, we must refrain from comparing its forma-
tion to that of dos, does. In the B. of An. occurs deupo,
20, 28, 29, 81 (God.); in the B. of Tal. ae deubu, 48. Duu,
An., 59; dyuu, B. of Hcrg., col. 810; dybi (fut.), col. 825; a
ffan rydyuu amser, 12. Gio., p. 119. Dyda rhymes with
o hona, gla, la, in B. of Herg., Shcnc, p. 304; dydo with
bro, bo, ffo, p. 305. Dedeuant etwaeth, dedeuho, dydeuho,
B. of Tcd., 10. Dedeuhar, ib., pp. 212, 213, a deponential
form, on which see Rcv. Cell., vi, p. 40 scq.

70. The other persons of aeth, doeth, gwnaeth are con-
jugated after the model of the perf. buum: euthum, later
euthym, etc. In modern times eis, dois, gwnais, eist, doist,
gwneist, etc, have been formed, " balbutientium puerorum
mera . . . barbaries" (!) according to Davies.

Cf. Cann y C; st, p. 585. Venedotian: mi yis, Yr Aru\, 17,
7, 56; gnyis, 13, 10, 56 (euthum ina, 17, 7, 56; mi ath, mi ddoth
etc). Powysian: mi eis ine, Cab. feiv. Tom. Dimetian: mi es,
C, iii, p. 525; pan ddes ati, p. 446; na'nese nw ddim on'd
wherthin, i, p. 332 (Pret. Sec); ond 'nest ti, iii, p. 624 (fe etho



ni, iii, p. 4 17; gnethe lii, i, p. 251; chi neithoch chi, a'netho nw, i,
p. 351; ib.: gna 3rd sing., os na 'nele, na 'nelset ti, iii, p. 184,

Eisym, deisym are combinations of eis and euthum;
cf. eisym, deisym, Y Cymmr., v, p. 166. Davies, Dict.:
ceniddum D.B. = cenais, cecini; this is from Marwnad
Dafydd vab Llywlyn by Dafydcl Benfras, 1240 {Myv., 2 p.
222). Sweet, p. 450, gives the secondary present doythwn,
gnaythwn, gnaythat, and gnt 1; etc. In this dialect also
cafael is partly influenced by doythwn, etc.; cf. keythat, etc.
(see 31).

71. Goruc (*ver + uc as in dwyn, amwyn, etc.) is said
byPughe(iu Coxe's Monmouthshire) to be a Gwentian form;
Hughes also, 1822, gives Silurian oryg North-\Velsh darfu.
In the Southern Middle-Welsh MSS. goruc is of extremely
frequent occurrence, though it is possible by counting the
iustances of goruc and gwnaeth that occur to establish
differences even in the various texts of the same MS. So
in the Mab. (1887), a wnaeth, etc, is rather more frequent
than a oruc in Pwyll pend. Dyfed and the following stories,
whilst in Gereint uab E., Per. ab Efr., Kulh. ac 01., a oruc
largely prevails. By comparing 32 of my Obs. on the Pron.
it will be seen that these rough results to some extent coincide
with the relative frequency, as therein stated, of the G went ia n
onaddimt and of ohonunt. E. Lhuyd, Arch. Br. (translation

1 Gwncuthur is often pleonastically used in Northern dialects. At
1 [olo Morganwg quotes, in Add. MS. 15,029, f. 114 (or 174a),from
Nb. 1 of F Gh'eal, blue wrapper: y gwna bawb wneyd eu goreu; and
remarks: "But a Northwallian can never sjteal or write without
these abominable ausliary yerbs" (to wbich also darfod belongs; see
g 57 i. Ib.: be calle na wnelont frwyscau, na wnelont ddyscu, etc, in
the Statute of Gruff. ab Cynan (also printed in Y Greal), "dull
Gwynedd". In )' Brython, L859, p. 208, l. ">". cerdd Imr ei gwneuthur
;i wnaeth occurs iu a jioem of Iolo Goch, to which the same writer,
E. Williams, remarks: "ef a wnaeth gwneuthur, a Northwallici.-in."


of the Gornish Tale), writes pan orygym dhacl (I went), and
pan orygsoch.

L. Cafael.

72. This verb (Corn. cafos, caffos, Bret. cafout, caffout)
occurs in Welsh as caffael, cafael, cael; and as gafael in
ym afael (Ir. gabail, *gabagli-), gafaelu, gafaelyd, and
ymaflyd, ymaelyd (besides ymgael). Davies, Gramm., gives
cf, cai, cei, caiff, cawn, cewch, cnt (and caffaf, ceffi, caffant),
caed, caffed, etc.; cawn, cait, cai, caem and ceym, etc.; caffwn,
caffem, ceffym, etc.; cefais, -aist cs, cst, cafodd, cdd, 1 cs
cafas, cawsom, etc.; ceusym, p. 259; cafwyf, caffwyf, pl.
cafom, caffom, caom, etc.; cafwyd, caffwyd, caed, cd, cafad
caffad; caid, ceffd, cawsid, etc. Thehistory of these different
forms is not clear; the first requisite is to collect the forms
existing in the more extensive Middle-Welsh MSS. I have
done this with regard to the Mao. (1887) and the texts
printed from Ll. Gw. Rh. (initial c and k are both written
as c in the following lists).

Cf. B. of Herg.: Ger. uab Erbin: caf (3 times), ceffy (11): cey
col. 807, ceiff (5), caffn (3), ceffch 784, caffant (2); caffyf (4),
ceffych 773, caffo (2); caffn 795: can 774, caffut 775, caffei (2),
ceffynt 807; sarhaet a geis 775: ceueist (5), cauas (6), cafas 779:
cassant (3); cassei 800; caffer (3); cahat 771; caffel (4): cael
773 (ymgauas 780, dyrchauel (4) ).

Per. ab Efr.: caffaf 685; ceffy (12), ceiff (3), caffn, cols. 662, 664,
caffant 700; caffyf (2), ymgaffyf (3), ceffych (2); can 671,
caffut (3), caffei (9); ceueis (3), ceueist (4), cefeist 695, cauas (7),
cafas (2), cassant 659; cassei 672; ceffir 683; cahat 669;
ceffit 695, 700: ceit 668 (dyrcheuit 698); caffel (10): cael (4)
(dyrchauel (3), drychauel 679).

1 Cdd occurs frequently in North-Welsh prints (18th cent., e.g., a
gdd, Trefriw, 1778, Y Traeth., 1886, p. 282; a gadd, Shrewsbury, 1763,
ib., p. 271; y cadd, Trefriw, 1798, ib., p. 424, etc.; and cf. ni chadd,
Y Cymro, Bangor, 1,1,1 849, etc. Gwdd (e.g., ac mi gafodd y tenantiaid
i gwadd ono i ginio, a dono Ue buon nhw . . . ., Yr Amserau, 25, 2,
1847) for gwahodd, to inyite, appears similar to cdd, from caodd?


Iarll. y F/ynn.: ceffy (2), caffn (2), ceffch 645; ceffych 630,
caffo 630; caffei 655, yingaffei 636; cefeis 633, cauas (2), cafas
(4), cassam 654, cassant 641; caffel (5); (dyrcbauat 636, dyrch-
afel 639).

Kulh. uc Olw.: caffaf 6, ceffy (17), caffy 827: cey 824, ti a gehy
826, ceiff 830; caffyf (3), ceffych (32, etc), caffo 811, caffom
821; caffut 844, caffei (3), ceffynt (2); ceueis (8), cauas (7),
cassant (2); caffat (4): cafat 843: cahat 818; ceffit 818;
caffer (3); ceffir (4); caffel (28, etc.): cael (3); (dyrcheuch,
dyrcliauei, dyrchauad; ymauael, ymauel, ymauaelad).

Bn "'/. Rhon.: ceffch 558, caffant (2); caffei (2); ceueis 559,
ceueist 559, cassant (2); (dyrchefit 566; dyrchauel 566).

Cyfr. Llud a Llcu.: ceffych (2); cauas (2); ceffit 706; caffel (2),
(ymauael 769).

Pwyll pt nd. Dyuet: caffaf (1), ceffy (4), ceiff (1), caffn 724,
ceffch 723; caffo718; caffn 718, caffei 723; ceueis (2), ceueist
(3), cauas (4), cassam 715; cawssei 724, caffel (3): cauael 714;
cael 715.

Brnw. rerch Llyr: ceiff (2), ceffch 734; caffo 732, caffom 734;
caei 728, 730, 732; ceueis (2), cauas (4), cassant 731; cassei (2);
cahat 728, 733; caffael 739: kynn kael o dyn yny ty gauael arna
736, cael (6); (dyrchauael 727, dyrcbauel 732).

Man. rah Llyr: ceffy (5); caffon 742; caffn (2), caffut 740,
caffei (1) caei (1); ceueis (5), cassont 742, -ant 743; ceffit (4);
(ydymeveil 745, 3rd sing.; ydymauelad 744, ymauael (3) ).

Math vb M.: can, lst pl., 753; cassant 753; ceffir 752: ceir
753; cael 752.

/,/. G'ir. llh.: Campeu Charl.: ceffy, p. 15; caffei 14; cauas 7;
caffat 6.

/;//. Tnrp.: caffaf 69: o chaf 54, ceffy (8), ceiff (3); caffwyf 51,
caffom (2), caffwynt 98; caffei (3), o chafyn nynheu .").">, ceffynt,
32; ceueis (5), ceueisti 68, cauas (13), cawssant 70 (ymgawssant (2) ),
cawsant 112; cawssei (ry-, 6); ceffir 108; caffat (4); cawssitll7;
caffel (ry-g., ymg., 16); (ymavaelaf 112, ymauaelawd 56; der-
cheueist, derchauel).

Bown o II.: caffaf (4): cahf 126, caf (3), ceffy (4): cey (4),
ceiff (2), (dyrcheif 1 "..">, cawn 182, ceffwch (2); cewch 128; caffo
1 16; cawn (4), caei (1), cai 140, 1 14, 154, ceynt (2); cefeist (3),
cauas (7), cafas (I), cawspam 127, cawssont 167, -ant (3),
cawsant(2); caws^ei (2), cawssynt 11:5; cawssoedei 170, caffat
L79, caliat 156; ceit 144; caffel (5): cael (6); (dyrchafyssant

Purd. Padr.: caffaf (1), ceffy (2); ceffych 199; caffut 199, caffei




(8): cai 193, ceffynt 190; cauas (2); caffit 199, ceffit 2C2 ■
caffel (2).

Buch. Meir Wyry: ceffy 214, ceiff (2), caffwn (1), caffant (1);
caffora 227; caffwn 213, caffei 227; ceueis (1), cefeis (1), ceueist
(2), cauas (6), cawssam 212, -ant 227; cawssoedat 224, kawssodyat
224; caffat 234, ny chat 222; ac yna y caffet ygkygor 218;
caffel (7); cael (3); (dercheuis 230).

From the other parts I only quote caffael, p. 241; cawesent,
p. 248; caffey, p. 262; caffadoed. p. 265; dyrcheif, p. 274; caffem,
pp. 289, 311; cael, pp. 311 (2), 314; cawsei, p. 310; na chyffit,
p. 322.

YS. Gr., p. 180: val y cahat.

Add. MS. 14.869: kein uyget am drefred dryfrwyd | kert gan
gyrt amgylch y allwyd | ceffid eu keinllith kwn kunllwyd | lceffynt
veryon yoreuwyd | lceffitur ymdwr am drwyd | heuelyt | Twrchteryt
y ar vwyd | caffawd beirt eu but yn yt wyd | keffid noeth noted
rac anwyd | lceffitor ym prufnad ym proffwyth | areith ym pryffwn
waedwyd, f. 56.

Salesbury, 2V. T.: pam a gaffat, f. 184; cahat, f. 2b. Huet: ny
chafad, f. 386/); ny chad, f. 39 b. Barddas, i: cafad, cad, cafwyd,
p. 32; caed, p. 64. Medd. Myddfai: cafwyd, p. 89.

73. The 3rd sing. of the s-Pret. is cafas, cafocld and cs,
cdd (given by Davies, Gramm., D. S. Evans, Llythyr.); cf.
ni chas, marg. chafodd, Gann. y G, p. 185; ond ni chas efe,
Iolo MSS., 28; a gadd, thrice on p. 22 of Lewis Dwnn's
Herald. Visit., vol. i; besides a gawodd, p. 34; a gavas, often;
Pass; a gad, p. 59 (cafes, Bardd., i, p. 238). Whilst dyrchafael
forms the plural of the s-Pret. in -assom or -yssom, etc,
cafael directly appends the characteristics of the s-Pret.,
cawssom, etc, being the result. Only in MS. Cleop. B5
have I found written sef cafsant, ff. a, 7b; na chafsant
f. 55&; a gaussawch, f. 92a; caussant, f. ha; n gaussant,
f. 16&; besides cawssant, f. ba (a drchauassawch, f. 186).
Cafesont is used by Sal., N. T., f. 159. Aw, from *af, is
treated like the old aw from *, and becomes in the later lan-
guage ow, o, yw; cf. my Beitr., 97-99 (Jes. Goll. MS. 141:
kowsant, kywsant; Add. MS. 14,986: a gosoch; Yr Arw.: mi
gywson i, mi gywsa (Middle-Welsh cawssei), cywsach chi),

VOL. IX. i



and ff< rald. Visit., i: a gowssant,pp. 22, 60, 90; a gowsant, p.
12 (2); Hope, 1765: ni chowse chwi, p. 7; Cdb. few. T.: mi
gowset ti, p. 39; os cowse fo, p. 15; na chowsan nhw, p. 39; etc.
In other dialects these forrns are replaced partly by cel- (see
74), partly by ceis, formed like eis, gwneis; cf. cs, Cann. y C;
ces i, Scr. C, ii,p. 6; cesest ti, p. 423; Cb. fcw. T: ni ches i,
cest ti, p. 6. G'eusym for cefais, like eisym, archesym (for
erchais), cersym and caresym, began to spread, according to
Davies, in his time.

74. In the Gwentian Add. MS. 14,921 (Maumdedlle's
Tratch) occur: ti gay, di gai, kyff, kn; kaffo, caffon;
Pr. Sec. y kay ef, ef a gay, e gae, y kaef (ef), f. 49Z>, kaffynt;
cafas, kawson; kafad, kfad; cael, cel, gaeffel, f. 22 (an
error similar to kaifail, MS. A, p. 68). F. 37 exhibits the
first example known to me of a flection common in the
modern South-Welsh dialects, viz., that based upon cel- as a
presumed verbal stem: y kelse. The Infin.cel (a chel) occurs
in Ll. Achau, p. 64.

Cf. Y Traeth., iii, p. 12, where pe celai e fyned, fel y celent, are
ascribed to the Gwentian dialects (for cai, caent). So in Y Tyw.
a'r (!.: ti gelset, ii, p. 66; Y Cymmr., iv (Iolo Morgauwg): ai celai
ef fyned i'r gwely; Y Fellten (Merthyr Tydfil): ni chelai, 1, 4,
1870; Y Bedyddiwr, viii, p. 4-4: celsai.

At Neath are used: c, cf, ci di, me geny, cfinw; ceso i,
cesosti, cesoti, fe gs; celwn i, celiti, cela fa, celyni, celyclii,
celynw; fe gespwd; i>.: gnethwd, gnespwd, fe wetwd, fe
g(y)merwd, fe welwd and fe welswd (a combinatiou of gwelwyd
aud gwelspwyd); cl (cjl, Pontypridd).

Bnt it is also written in texts intended to represent the
colloquial Language in tlie Dimetian parts of South-Wales.

Cf. Seren Cymru (Carmarthen): 'chelwn I ddim, i, p. 374; chele
ui ililiiu degwn na treth eglws, iii, p. 5; chele 'r ddou grwt 'na
ddim o p. 227; ond chelswn I ddim, p. 226 ctc.
Seren Oomer,
xxx\ i: ia chele fe, fe alle y cele dyn da gam gyda ni, p. 362;
a chele, ac fe gele, p. 157.

In Add. MS. L4,921 ymaelyd also occurs: ar tan yn



dechre ymaelyd, heb yr tan gael ymylyd ac hwynt; cf.
B. of Hcrrj.: ymauael, col. 824; yn ymgael ac ystlys y
llannerch, col. 700; o yingael ar gr a dywedy di, col. 711.

M. Ox various Verbs, mostly Defective.
(Zeuss, Gramm. Cclt., 2 p. 604-6.)

75. On cigleu, ciglef, see Rhys, Rev. Cclt., vi, p. 24; cf.
B. of Herg.: ry giglef, ny chiglef i, a giglef, cols. G74, 800,
801, 835, 836; ny chigleu i, col. 834; 3rd sing. kicleu, col.
780. Add. MS. 14,869: y clywspwyd, f. 240. For examples
of clowed and clwed for clywed, clohod for clyhod, see my
Bcitr., p. 43. Cf. clwas, clwasti, clyw, ni glwson, Inf. clwad, in
Neath; North-Welsh: pen glwan nhw, Yr Ams., 4, 11, 1847;
mi glwis i, 1, 7, 1817; clwad (Imp.), 15, 1, 1851, etc. *Clev-,
*clov- became *clow (pre-tonic clyw) and *clou-, cleu, perhaps
also clo- (clo-bod, cly-bod), depending probahly on the quality
of the following vowel; cf. taraw, tereu, etc. (Q6). In the
South-Welsh MS. of Han. Gr. cib C. cigleu is used; in the
North-Welsh MS. clybu, cly wodd, Myv. Arci., 2 pp. 723, 725,
726, etc. 1 Davies, Dict., gives degle, ehodum, heus, ausculta
(Spurrell has degly v., to hearken, to listen), probably

1 In Seren Gomer, 1814, p. 19, clywed peth drewedig, clywed blinder,
clywed bwydydd blasus are given from South-Welsh dialects, clywed
being used there for arogli (L. Morris, Adrl. MS. 14,923, f. 1346,
mentions arogl (scent, smell), brŵd (hot, warm), rhawg (for a long
while), fo, gan, efo, as words not occurring in South-Welsh dialects).
Teimlo, archwaethu: In the Camhrian Journal, 1856, p. 248, is given:
beth yw 'r blas cas rwy'n archwaethu ar y cig y-ma? in Dyfed,
'ndeimlio iu West Glamorganshire, ^n glywed in East Glamorganshire
and Monmouthshire; clywed ('' to taste" and ' to smell"), in Monmouth-
thire and East Glamorganshire, but never in AYest Glamorganshire. D.
S. Evans, however, saysthat clywed is also used in Dyved in that sense.
Add. MS. 15,049, f. 22b: yn ylle i clowo y klaf y dolur. To this cor-
responds what is said in vol. ii of the led Dragon (pp. 38-40), Cardiff,
1882 et seq., that Welshmen speahing English say "to hear a smell'';
also " to sing a piano''; cf. canu telyn.

i 2



*de-gleu; cf. Daf. ab G\v., Poems: degle ferch, p: 112; degle
'n nes, p. 218.

76. Davies, Dict., has *handid, idem quod hanfydded,
sit, existat; *handoedd, pro hanoedd, fuit, erat; *handym =
y.lym, sumus; *handyfydd, pro hanffydd, erit. Examples of
handid are frequent in Middle-Welsh prose; the other forms
occur in the older poets; cf. Add. MS. 14,SG9 (Gogynfeirdd):
Cyd vuam gyd ac ef handym oll gyuaclef handid tegach
teulu nef, f. 193&; heueis dwy handid mwy eu molawd, f.
234a; handid, f. 70&; handwyf, f. 745; handoet eu hachoet
kyn eu hechig, f. 85.
Eleven forms occur in the poem
of Llywelyn Bardd y Lywelyn vab Ioruerth, f. 1165 (printed
in Myv. Arch., i, p. 358 = 2 p. 247): handid (twice), handwyf
(twice), handwyd, handyuyt, handym, handoetud, handoet
(Iwice); " handythuagwyd peuyr yn penn erchwys yn oreu
keneu kynon vegys."

77. On dyre, colloq. dere, see Ehŷs, Rcv. Cclt., vi, p. 26.
B. of Ilcrg.: na dyret ti, col. 794; dyret, cols. 776, 799, 800,
801; dyret, Ll. Giu. Rh., pp. 120, 173, 187. Deret gyt ami,
li. Cct., viii, p. 9; dyret, p. 21. L. Morris, Add. MS. 14,923,
f. 133a, has.: South-Welsh dere 'n gloi North-Welsh tyr'd
yn gwit, sydyn, fuan (come quickly); he also gives tyr'd
from Anglesey (Add. MS. 14,944, f. 153&); cf. tyd, Sweet,
p. 420. Richards, Dict. (1753): South-Welsh dyre, dere
NTorth-Welsh dyred, tyred. Dera gyta f, Neath; dera geno
i. Aberdare.

78. Zeuss, 2 p. 60(1, merely mentions liwde (ecce, accijic,
sume). Cf. Ll. Gfw. llh.: a hwdy ditheu ef, p. 52; hwdy vy
ffyd, p. 5 1; hwde, ]. 179. Davies, Gramm., gives hwde, liwre,
accipe, pl. hwdiwch, hwriwch, \\\\ rewch. Rowlands, Gramm.,
*107, and Williams (Dosparth Edeyrr) give North-Welsh
hwde, South-Welsh hwre; Williams has also hwda, hwra
(cf. )' Brython, 1861, p. 112: hwda, from Denbighshire).
The etymology of these words is obscure. The only cxi>la-



nation of them not altogether improbable is, in my opinion,
that hwre stands for *wre, *wyre, on which see Rev. Celt.,
vi, p. 28 seq. Calliug to mind Corn. wette, wetta, otte (Zeuss,
2 p. 006) from *wel-te, ancl the Welsh tyd, from tyred, ex-
hibiting phonetic changes due to that advanced degree of
phonetic evolution to which isolated words like these inter-
jectionally-used imperatives have attained, I would suggest
further that hwde contains hwre with the pron. of the 2nd
pers sing. It would be more satisfactory, of course, if hwre
could be proved to contain a deponential ending. Or is it
merely a phonetic change, like Breton hirio: W. heddyw?

79. Dabre is not mentioned by Zeuss. Spurrell, Dict.,
has dabre, come, come hither, and dabred, dabredu, dabrediad,
which latter words I do not remember to have met any-
where. Cf. B. of Herg.: dabre, col. 717. Ll. Gio. Rh.: dabre
di bellach, p. 72; dabre yr maes, p. 159; na dabre di, p. 54,
etc.; Y S. Gr., pp. 160, 167, 169, etc.; Add. MS. 14,969: uy
adas ny debre, f. 148 {Gwalchmai i Efa ei wraig); Daf. ab
Gw., Poems: debre 'r nos ger llaw 'r rhosydd dan frig y
goedwig a'r gwŷdd, p. 31, also p. 134; Salesbury, Dictionary:
debre, harken; JY. Test.: dyred, marg. does, dabre, f. 41;
dabre ac eclrych, marg. dyred a' gwyl, f. 379 (Huet); tyret
ac ■ edrych in Morgan's eclition (1588). Davies, Dict., gives
also anebre, marked as obsolete: " vid. an an-nebre, ab an et

80. Moes, pl. moeswch, da, praebe, age; moeswn, moesant
" nonnulli dicunt", Davies, Gramm.; cf. moes vy march
B. of Hcrg., col. 717; moysswch y llythyr am march am
cleclyf im, Ll. Gw. Eli., p. 131. It stands probably for ym
oes, and if moeswch is a neo-formation like doswch, it would
be formed similarly to does, and allow us to assume, if neces-
sary, an older *ym os (see 69)?

81. The deponent verb hebr has been pointed out by
Pihŷs in the ordiuary orthographies heb yr (Y Gymmr., viii,



p. 161, with reference to page 114). Cf., c.ej., Gnocl. Bardd
Cwsc, 1759: ebr un, ebun arall, ebr finneu, ebr ef, ebr y
trydydd, pp. 4, 5, 6. L. Morris, Add. MS. 14,92:3, f. 134,
gives South-Welsh rnenta fe North-Welsh meddai fo; also
D. S. Evans, Llythyr.: Soutli-Welsh, myntai = ebai, raeddai;
cf. mynta fi, Y Bcd., viii, p. 100; mynta finau, ib. (Mon-

Septembcr 30, 1887.


In my Bcitr., pp. 78-9, and Obs. on the Pron., Y C, viii,
pp. 157-9, 1 drew attention to cynnag pwy, etc, occurring in
the lOth cent. Add. MS. 14,921, and in the dictionaries
(Davies, Pichards, Pughe, Spurrell), but not, as far as I was
then aware, occurring in other MSS. or mentioned in books,
etc. I have lately loohed through the first ten volumes of
Yr Amserau (Liverpool, 1843-1852, fol.), and read Llythyra/u
'rhen Ffarmwr, in which, by the way, few dialectal forms or
words occur which are not known also from Cab.few. Tonios,
the Letters of Wmffra 'rgwŷdd in Yr Arweinydd, a plagiarism
of the " Old Farmcr's" letters, etc. Before these letters begin to
make a regular appearance two evidently fictitious letters of
an opponent are inserted, written in a Sonth-Welsh dialect,
or, rather, containing idiomatic forms of several Southern
dialects. In these, ganta pvr'n, ganta p'in, for bynnag pwy
uii, }>a im, occur; cf. darlleniais yn eich papẁ wythnos i
hedd-i l/'th/r oddiwrth ^Hn Ffannwr" yn gweid yn erbio
dẁg ganta pwi'n iw e, mai yn d;i////-//<>s nad ois dim mohenofe
gwedi ca ond riw gettin bir o addẁg. Mai yr hn ŵr (n</
weddw ganta p'ir), yu gwrthddeid ei hin yn ofnadw; ....
ond ganta pwi 'n iw e, Yr Ams., 17, 12, 1840. Shortly after



reading these letters I was told 1 that gynta p'n, gynta beth
were commonly used in Glamorgan, and also 'ta pn, ta
bcth (so at Neath), the first syllable of gynta, which is less
strongly accented than p'un or beth, being dropped. Cf.
'dos dim tu fas ta bcth, falla fod llawer tu fewn, Y Fdltcn,
(Merthyr Tydfil), 1, 4, 1870 (Glyn Ebbwy). The connection
of these forms with cynnag pwy, etc, is obvious, but it
remains an open cpiestion, worthy of further consideration,
whether cynnag is an old form and cynta a modern etymolo-
gising transformation of it ('ta beth, for beth bynnag in North
Wales, meaning " the frst thing" in the sense of the German
"das erste beste ding",for "whatsoever"), or whether cynnag
is a combination of bynna(g) and cynta(f). In Add. MS.
14,921 cyntaf is twice written cyna; cf. ac or il hono y ddir
y siprys lle my llaweredd o winwydd. yn gyn ymynt yn
gochon a chwedy hyny yn wynon yrhan fwya, f. 5a; yno y
dwad yn harglwydd gyn wrth samywel, f. 30. I have not
noticed such orthographies elsewhere, and they are possibly
caused by cynt being pronounced cyn.

1 I received this and other information on the dialect of the Neath
Valley, quoted in part in the preceding pages, from my friend Mr. S.
Mainwaring, a native of that part of Wales.

January 31, 1888.


To footnote on p. 75: lladd and lleas. Lladd, "to cut" or " kill", is
for slad, as in the Irish saidim, of approximately the same meaning,
while lleas is a derivative (like priodas, cymdeithas, galanas, etc.) from
lle for seg or slig, as in the Irish sligim, " I slay".

To p. 101, line 19: cnbyddiaeth. Cnabyddiaeth or cynabyddiaeth is
for cydnabyddiaeth, like prynhaum or pyrnhawn (better prnhaicn), for
prydnawn, always accentuated on the nhawn; compare also crynu, "to
trernble", for *crydnu.






[1.] The study of Brythonic declension made very slow
progress until it was discovered that certain plural termina-
tions are in reality the suffixes of collective nouns, and that
j under certain conditions of accent is liable to become d, as in
Greek (Rev. Cclt., ii, pp. 115 et seq.). Up to that time its results
had been nearly limited to some very obvious identifications
of plural endings, and the recognition of a few oblique cases
in adverbial formula?. Even since then the phenomena of
Brythonic declension have seldom been regarded from any
other point of view than the possibility of their throwing some
light on the more carefully studied grammar of Irish. It is
true that the materials afforded by existing dictionaries and
grammars are quite insuffcient for a history of any of the
plural terminations, and that each individual word nmst in
consequence be followed through the older stages of Welsh,
Cornish, and Breton before any opinion can be formed upon
it. Analogy was at work at a very early period, and in many
later; and in a large number of cases has given the same word
several plural forms, creating a difficulty which has been
further enhanced by the position which different lexicogra-
phers have taken up with regard to these matters. The stems
of a number of nouns have been ascertained with tolerable
certainty, but nothing final can be undertahen without the
publication of trustworthy editions of a much larger quantity
of Middle-Welsh texts than is available at present, and the


preparation of a dictionary of the moclern language which
shall comprise the dialectal forms. I shall confine myself here
to drawing attention to some little-noticed facts which may
be at least stated, though at present not always explained.

[2.] Lewis Morris, in a letter of 1762 (YCymmr., ii, p. 157),
says in his quaint way, " If South Wales men had wrote
grammar, we should have proper plural terminations instead
of -au, etc, etc, and abundauces of liceuces of the like kind."
The first two editions of Owen Pughe's Dict. are criticised in
Y Brython, i, pp. 19-20, and also by the editor of the third
edition (Pref., p. xi), on the score of his too frequent use of
-on ancl -ion instead of -au, even where -au is in general
use. He had a preclilection for the apparently more archaic
-ion in place of the common -au. By this, as by similar
fancies, his work is deprived of much of its valuc.

Cf., e.g., L. Giv. Rh.: oc eu gweieu, p. 104; o dyrnodeu,
a bonclusteu, a chwympeu, a gweioed, a chellweir, p. 109.

Y Drych Chr.: archolhieu, archolhiou, f. 27&.

Davies, Dict.: archoll, pl. erchyll; on which L. Morris (Add.
MS. 14,944, f. 26a) remarhs: "The coinmon plural is archollion."
The word is not used in Glamorganshire, where they employ clwyf.

Sp., Dict. 3: nant, nentydd, nannau; L. Morris: neint, nentydd
(Add. MS. 14,909, f. 50).

E. Lhuyd: asen, eisen plur.: N. W. asenna?, S. W. ais (Arcli.
Brit., s. v. ' Costa' 1 ); but asan, 'sena, are the Neath forms too.

Rowland: llythyr; N. W. -au, S. W. -on (Gratnm. 4 , p.29); etc.

[3.] Monosyllables like march form their plural by the so-
called internal i, as meirch. This kind of formation was also
used in words of more than one syllable, as in aradr, plur.
ereidir (B. of Carm., No. ix 2 ). In these words it was sup-
planted at a later period by the apparent change of a into y,
as in erydr. Davies, in his Gramm. (pp. 36-38), gives:
erydr, old ereidr; old peleidr, rheeidr; becbgyn as the
plural of bachgen [Glam. bachcan, bechcyn]; llenncirch
(ToloG.)and llennyrch of llannerch; ieirch (Iolo G.) and
' P. 52, col. 1. i Shene, vol. ii, p. 11; MS., f. 17.


iyrchod of iwrch [cf. lewot a iyrch, Ll. Gw. Rh,, p. 74, cf. p.
76; ierch, p. 143, froni ieirch]; ceraint, also cerynt, of carant;
Id., Dict.: cleifr of dwfr; emyth and emeith of amaeth;
heiyrn and heieirn of haiarn (cf. Sp., Gramm. 3 , 28; in
Glamorgan dwr, dyfrdd).

Here we see the old groups a, plur. ei, a a, plur. e ei,
and a c, plur. e y, as well as o,w, plnr.?/(mollt myllt, llwdn
trillydn, canllydn, Davies, Gramm,, p. 35), completely mixed
up. The general tendency seems to he to extend y over the
dissyllables that contain a a, except some like dafad, carant;
while the opposite extension of ei over both mono- and dis-
syllables containing a e, o, w is much more limited and pro-
bably does not now obtain. There are some plural formations
existing of words that do not evidently differ in structure
from all other regular words, which remain unexplained, and
prevent our coming to a really fair view of the neo-for-
mations just quoted I mean the North-Welsh plurals, ifinc,
llygid, bychin of ifanc, llygad, bychan. Ab Ithel (Dosp.
Ed,, 1571) gives llygid from Denbigh, Flint, and Meri-
oneth (also bysidd for bysedd, from Denbigh and Flint). I
collected in Bcitr., 92, other North-Welsh instances of i
for ei, and in Y Cymmr., ix, pp. 67-8, I considered North-
Welsh collis, ceris for colleis, cereis from the same standpoint.
I have since heard thatthe same forms are used in Glamorgan;
cf. ifanc, plur. ifinc (jenctyd, jangach, janga); llycacl, pl. llycid;
bychan, pl. bychin; bys, pl. bysidd (but e.g., gwracacld); also
merchid; pryfetyn, plur. pryfid; arath, plur. erith (erill). In
the CambrianJournal,Y,\>\). 208-9, 1 find the plurals offeirid,
merchid, bychin, gwinid ( = gweiniaid),from GwentandMor-
ganwg. In the same way rhois is used both there and in North
Wales, the regular alteration of Mid.-W. -ci in final syllables in
Gwent andMorganwgbeinginto a, just asfinal e becomes a. In
fact, ei first became e,as in the Dimetian and Powysian dialects,
and later all e's were made into a. I have found also rhois, and


this form only, in mucli older texts than those given in Y
Gyminr., I. c, and therefore I consider collis, ceris, etc, as imita-
tions of the very old and unexplained type rhois. The plurals
with i are also as yet unexplained. I should add that
merchid, prylid cannot be compared to bysidd on thescore of
a change of e into i, though the literary forms are merched,
pryfed. They are identical in formation with offeirid and ifinc,
for amongst others William Morris (Add. MS. 14,947, f. 39)
gives merchaid, pryfaid from Anglesey, where these forms
are still in use. Cf. also Hanesion o'r Hen Oesoedd, 1762:
mercheid, pp. 28, 90, etc.; merchad, p. 72 (Carnarvonshire);
P. C, No. 28: merchaid (Anglesey); Yr Arw.: merchaid,
Feb. 12, 1857; Jan. 23, 1859, etc.

Davies, Gramm., p. 37, gives from the works of the old
poets e i, as well as the common e y, as a plural form from
the singular a e, this being proved by rhymes; e.g., cessig,
cerrig, menig; tefill, pedill; llewis, etc. The explanation of
these ibrms may be connected with that of the plurals given

[4.] Daint, daigr, saint, plurals of dant, dagr, sant, are used
in N.-W. dialects as singulars. Salesbury, 1567 (Ellis, Early
Engl. Pron., p. 747), remarlcs that for epestyl, caith, daint,
maip, saint, tait, in his time apostolion, apostolieit, caethion,
dannedd or dannedde, maibion, santie or seinie were beeinning
to be used, and that in N. W. daint, taid (fathcr, not taid,
teidion, grandfathcr), were used as singulars. Davies, Did.,
has "dant, pl. daint, quae vox apud Yenedotas pro sing. passim
usurpatur." L. Morris (Add. MS. 14,909, f. 756) gives N.-W.
saint for sant; D. S. Evans, Ltyth., s. r., a sing. and plur.
deigr. [f only deigr and daint had been preserved, we
might be inclined to consider them as old collective nouns;
lmt saint cannot be such, and so the most probable explanation
a' them is that they arose at the time wlien the plural
character <>f tho older plurals, f'ormed by what we now seo in


its result as an internal change, mab *meib, dagr *deigr, was
accentuated by the addition of the constantly extending -au or
-on, -iau or -ion, so as to yield meib-ion, *deigr-au, after the
analogy of dagr, dagrau. We may suppose that at this time
deigr. etc, were erroneously abstracted as singulars from
these new plurals, and came into general use in this capacity
in certain parts, in this particular case in North Wales.

[5.] The plural termination -oedd is commonly assigned
to the stems in i, but it has not been successfnlly explained.
It would be the outcome of an older *-jes, which may have
been the result of analogical operations similar to those
which produced the Greek forms like *7ro\r)je<;. We must
not consider this as the only ending of the i-stems, for -ydd
from *-ijes, and -edd from *-ejes, are also early endings of
t-stems, coinciding in sound with the results of thejo-stems,
and with collectives in *-ej. As -i and -ydd are phoneti-
cally equivalent, and outcomes of the same older j- forms
with differing accent, we may assume i also as a termination
of the t-stems, either original or due to analogy; and we need
not wonder at the analogical extension of these endings of
the -stems, considering how they partly coincide with the
jo- stems and the collective nouns, and also how far -au,
the ending of the much rarer w-stems, and -on, that of
the i-stems, were extended.

This view of the i-stems is corroborated by a number of
words which form different plurals, in -oedd, -ydd, -edd, and -i.

Cf. trefi, trefydd, sometimes trefoedd, L. Morris. Add. MS.
14,934, f. 16.

Dinas, gwlad, ynys, tir, caer, myiiydd. Add. MS. 19,709:
mynyded, f. 57; y kestyll ar kaeroed ar dinassoed, f. 39a;
keyryd, f. 396.

Ll. Gw. Rh.: dinassoed, dinessyd, p. 22; dinessyd ar keiryd,
p. 74; y dinassoedd ae geiryd, p. 74; dinessed, p. 281; keyrod,
p. 24, kayroed, p. 33; ceiryd, pp. 45, 48; ceuryd, p. 54; ceyryd.
p. 21; gwledi. p. 187; lluoed, p. 22; lluyd, p. 21.


Sal., N. T.: dyfroedd, dyfredd, f. 382?;; mynyddedd, f. 3806 j
mynyddey, f. 380i; b\vystviledd, etc.; gweithrededd, f. 157b
Add. MS. 14,986, f. 37/,: tyredd.

Ynys, llys, tir, llawer. -oedd, poet. -edd, Davies, Gramm., p. 30.

The Middle-Welsh examples could be greatly multiplied. In
most of these words -oedd predominates, more rarely -ydd.

-oedd is pronounced -oedd or -dd; c.g., S.-W. gwithredodd,
Add. MS. 14,973, f. 102; N.-W. cantodd, mylldyrodd, Yr
Arw., May 28, 1857; byth, bythodd, etc. It may be men-
tioned that -ioedd, which is not given by Zeuss, also oecurs,
although cominonly only in milioedd. Cf. llawer o villtiryoed,
B. ofHerg., col. 1112; eithauioed freinh, MS. Cl. B 5, f. 19
(also ff. 61&, 62); eithaueoed, f. 97; tiryoed, L. Gw. Bh.,
p. 165; iethioedd, Sal., N. T, Giccl. L, f. 3846; brenhinioedd,
Add. MS. 14,916, ff. 36, 366; several instances in the
Gwentian Add. MS. 14,921; Hanesion or Hcn Oesocdd: ieith-
ioedd, p. 59 (twice; ieithia, ib., in rhyme); etc.

[6.] In N.-W. dialects -oedd has been partially ousted by
an apparent ending -fŷdd, the plural of -fa, from *magos, con-
tracted from *-fe-ydd, according to the law which obtains in
gwŷdd, for gweydd, ' weaver', and Llŷn for Lleyn, which has
lost a *g between the two vowels.

Torfŷdd, porfŷdd occur in Anglesey and Carnarvonshire, and
in imitation of themllefŷddfor lleoedd. See Caledfryn's Gramm. 2 ,
p. 59; and cf. rhai llefydd yn y Merica, S. C, i, p. 372 (Merioneth);
llefudd, Yr Arw., Feb. 2, 1859; Sweet, p. 429, etc. It is not clear
to me why E. Lhuyd, in At y Cymry, Arch. Brit., says: "amryu
levy5 jy Ildnry", p. *1, as he otherwise uses S.-W. (Dimetian)
forms. On deuwedd see my Beitr., 80, n. 33.

[7.] The early consonant-plural *chwior took later the col-
lective endings -ydd and -edd. Cf. dw chwored, MS.
Tit. D 2, f. 177; chwiored, 7r,p. 100; chwiorydd and
-edd, Add. MS. 31,059, f. b; in Neath, whr, whiorydd (ib
gwrac-add, modrypodd, commonly "modrybedd.in N.-W. mod-
rabedd", Richards' Dic.). Chwaer represents *svesr-, if caer
represents *-casr-: or else *chwaear (cf. gwaeanwyn) became


in its older form chwaer. Chwa-er from *sve(s)er- is a third
possibility. Chwaer had its old *ves treated like gwaeanwyn,
O.-W. guiannuin, containing *vesant-. Though guiannuin 1
(in Merioneth gwinwyn, Ehŷs 2 ) and chwiorydd exist, we have
gwaeanwyn and chwaer, containing *wa, *va, from *vo for *ve;
but it is not clear for what reason the syllable *ves became
either *vi(s) or *vo(s), wa(s), although it is true the next
syllable contains different vowels in both cases; chwaer from
*svo(s)er, *sve(s)er, chwior- from *svi(s)or-, *sve(s)or-. Chwi-
orydd as an isolated form was altered by analogy; cf. chwae-
ored, MS. C. B 5, f. 249, col. 2 (Dares Phrygius); brodur
a chwayorydd, Add. MS. 14,987, f. 356 {Araith y Trwstan).
In The Bed Dragon, ii, p. 420, chwaeriorydd is men-
tioned. This form is actually printed in Llyfr Achau (Her.
Vis., ii, p. 12) as chwaerorydd. Pughe gives chweirydd, pro-
bably to be considered like meusydd, pl. of maes.

[8.] The ending -awr, later -or, is almost completely lost in
later Welsh; it is frequent with certain nouns, as llafn,
byddin, ysgwydd, gwaew, etc, in the older poems, those
printed by Skene and in the Myv. Arch., but it harclly ever
occurs in Middle W. prose texts (cf., however, yr ieuawr, Ll.
Gho. fih.,]). 7). From O.-W. cf. poulloraur (*pugillr- + -r-),
M. Gap? In Breton this ending is still common as -er, -ier;
cf. er mneier in aTrc. poem, Rev. de Bret., lst ser., v, p. 408
(mene'io, 4, p. 170), etc, in W. mynyddoedd. The Welsh
ending is mentioned in Dosp. Ed., 489, omitted by Zeuss,
but exemplified in E. Evans' Stucl. n Cymr. Phil., 12. 4
On its Irish equivalents opinions have been advanced in Thrce
Micl.-Ir. Hom., p. 135, and Togail Troi (LL.), Introcl.

It is possible that -or has survived in certain plurals for
which a new singular was made by means of -yn or -en in

1 Ovid Glosses, 40&; Stokts, Cambrica, p. 236. 2 Lcct. 2 , p. 27.

3 Capella Glosses, No. 28 (rch. Carrib., 4th Ser., iv, p. 7).
* lb., p. 141.



marwor, marwar, marfor, ' embers', for instauce but as a
distinct plural terminatiou it occurs in later Welsh ouly in
gwaew, gwaewawr, whence *gwaewor, gwaewar, like gwat-
wor, gwatwar, niarwar, chwaer, etc,

Cf. gayawar (sic), B. of Ta., xliv (Sk., ii, 199); gwaywar, Ll.
Gw. Rh., pp. 59, 67, 69; gaewar, B. of Uerg., cols. 698 (R. B.
Mab., p. 82), 1093, etc.

Even here it is replaced by -yr; cf. gostng gaewyr, B. of
Herg., col. 1093; y gweiwyr, Wms., Hgt. MSS., ii, p. 304;
gwewyr, Y S. Gr., p. 230 (gwaywyr et vulgo gwewyr, Davies, Dict.).
This -yr can only be the ending -yr (-r) of brodyr, pl. of brawd,
analogically transferred. Besides brodyr, broder less ofteu occurs,
rhyming with full syllables in -er (passives, etc.) in poets, as in
L. Glyn Cothi, pp. 42-3 (see p. 269, infra), etc.

[9.] Tai and lloi are older S.-W. plurals, commonly used
in S. W., but replaced by teiau, lloiau in N. W. Cf. Hughes,
1823, p. 29; Y Traeth., iii, p. 9; D. S. Evans, Llyth.ỳs. w.;etc.

Ll. Gw. Rh., pp. 94, 97, 110: Dagr-eu-oed; Add. MS.
19,709, f. 58&, etc.: blod-eu-oed, llys-eu-oed.
Cl. B 5, f. 102&. These forms should be compared with Breton
ones like bot-o-ier, bot-o-io, Ernault, De l'urgmce, etc, p. 14;
the additional plural-ending seems to convey a more collective

Cf. also or dyededigyon llysseue hynny, Add. MS. 14,912,
f. lla; o oll dyededigyon llyssyoedd hynny, f. 79 (y for eu, S.-W.);
llysewn, Uisiav, Add. MS. 15,049 (17th cent.), ff. Aa, 20a;
llissewyn, B. of Herg., col. 436. In Neath, llysewyn, pl. llysa
(il>., blotyn, pl. blota).

On eu: ew, cf. gneuthur creu yr moch, B. of Herg., col. 754;
y creu, col. 766 (R. B. Mab., p. 63, 1. 3; 78, 11. 8, 11, 16): crewyn;
gieu: giewyn; ceneu: cenawon, etc.

On the words meaning ' day' see Rhŷs, Hibb. Lect., pp. 116-118.
Cf. die6-ed, Add. MSS. 19,709, f. 2hb ■ 22,356, f. V2a; diewoed,
Cl. B 5, f. 235/;, col. 1; die-oed, f. 2Ub, col. 2; dieuoeddin Davies'
Dict.; ew: eu, as in the words above given.

A plural ending is also often incorporated in singulatives
formed from English loan-words; cf. sklait-s, sing. sklait-s-au
('slate'), tatws, sing. tsan, etc, Sweet, p. 437. On Breton
analogies see Rev. Celt, vi, pp. 388-9.


It must not be assumed tliat the suffixes -en and -an
are of the same phorjetic value in all words which exhibit
them in the present state of the language. In a few such
words -aen is the older form of the suffx, as is confrmed by
other Brythonic languages. So agalen, ogalen: agalayn, ' cos',
MS. Vesp. E 11 of the Latin Laws (Owen, p. 851). Cf. 0.-
Corn. ocoluin, gl. ' cos' (Stokes, Gambr., p. 241); llamhystaen
( = llymysten); see Zeuss, C C 2 ,p. 291. Cf. also croessaeniait,
MS. L, p. 182 ( = croesan, -iaid); maharaen, MS. A, p. 135,
pl. meheryn, p. 160; maharen, MS. K, pl. meherein, MSS.
C, D, B, E; in the Latin Laws: maharayn, Hgt. MS., p. 791;
maharain, Harl. MS., p. 862; maharaen, MSS. Calig. A 13,
f. 183&, Tit. D 2, f. 536. Cf. O.-W. maharuin {bis), B. of
St. Cliacl, pp. 18, 19. 1 (Lib. Lancl., pp. 272-3, nos. [3]
and [4].)

This reduction of the fnal unstressed syllable may be com-
pared to that in gallel: gallael, cafel: cafael, gadel: gadael, etc,
but the details of the process are not all clear. Is maharuin
an error for *maharain, or perhaps (cf. ocoluin) the representa-
tive of a later *maharwyn, like mollwyn, -od (which might be
an imitation of it), gwanwyn, etc.? In other words have we
again -an and -ain? Cf. rhiain and rhian, adain and adan
(also aden), celain and celan, all fem., and all forming plurals
in -edd and -ydd. These are stems in - in which both the

1 Myharan, myheryn, 'wether, ram,' Sp., Dict. 3 L. Morris, Add. MS.
14,923, f. 134, says: S.-W. cig maharen = N.-W. cig mollt (Bret.
maout, Ir. molt). Richards, Dlct.: " maharen, in N. W. and in some
parts of S. W. as Glam., a ram; in Powis and in the greatest part of
S. W., . . . a weather"; Arch. Brit.: Dimet., Powis., 'vervex'; Vene-
dot., ' aries', p. 285, col. 2, s. v. 'sheep'; Richards, Dict., s.v. yspawd,
' a shoulder': N. W. 'Spawd mllt. 'Spold gweddar, S. W. [Dimet.].
Palfais gweddar in Monm. Glam. and Prec.
Y Gwyl., vi (1828),
p. 207: S.-W. hwrdd = N.-W. myharan.
Palfais, in Neath palfish,
is the shoulder-blade. In Neath, my?haran is a ram, given also in
the Cambr. Journ., iv, p. 208 (minharan = hwrdd, fem. dafad), from
Monm. and Glam. Cf. Sanskr. msha; mynharan contains myn, 'hid',
introduced by popular etymology.

T 2


nominative aected by the termination *-i and the other cases
not affected by it have been generalised. Ir. anner (fein.) is
represented in Welsh by anner, anneir, ' heifer'; ' bucula,
junix', Davies, Dict. 1 (anner, anneirod, Sp., DictJ).

[10.] So far these notes on nominal declension; to which
some details on dialectal difierences in the gender of nouns,
in nouns of relationship, etc, may be appended.

As rnasc. in N.-W. and fem. in S.-W. I find mentioned: troed,
effaith, ysgrif, rhif, nifer, clust, sain, munyd, man, golwg, ystyr,
gradd (Dosp. Ed., 471; dwy droed, Williams Pant y Celyn, Y
Traeth., 1870, etc); hanes, ciniaw, gwniadur, cyflog, cld, clorian,
gr, Rowland, Gramm.*, p. 39; N.-W. pellen y pen glin, S.-W.
padell y benln, D. S. Evans, Lhjth., s.v. 'penln'; alarch, /(/.,
WeM Dict.

Braicg, fem., is, according to Rhŷs, Loamcords, s. v. ' brachium', 2 a
masc. in Salesbury's language and still so in Carnarvonshire as the
spur of a mountain. The same author has recorded the older comm.
gender of dyn and the fem. gender of haul (still preserved in some
parts, t./7..about Ystradmeurigin Cardiganshire, Rev. Celt.,\i, p. 40),
and drawn certain conclusions from them (Hibb. Lect., pp. 92, n. 1,
572, n. 2).

Cwpan, perjnill, pontbren, canwyllbren, canrif, clust are given
by Rowland, l. c, as masc. in S.-W. and fem. in N.-W.

1 According to Y Geninen, iii, p. 19, Glamorgansh. y dreisiad (in
Neath trishad)=N.-W f . yraner; also heffar (' heifer'), YrArw., Feb. 20,
1859. Anner is the Dimetian word, C'ambr. Journ., iii, p. 252. Treisiad
is also the S.-W. word for N.-W. bustach, ' steer' (Davies, D'ui.). L.
Morris, Add. MS. 14,944, f . 28: llo dyniewed bustach ych are the
successiye yearly naraes used in Anglesey. Enderic, Juv. Gl. (8tokes,
Cam/r., p. 2UG); W. Lleyn's Vocab.: enderig = bustach. Dyniewed,
dyniawed, pl. dyniewyd (dinewyt, Z.-, p. 282), ' steer, heifer', Sp., Dict. 3
It is the old Cornish deneuoit, 'juvencus', Corn. Vocab., f. 9a. It
seema to have been submitted to popular ctyinology, if the following
notice, taken from Y Cylchgr. (Abertawy), 1853, p. 17, be genuine:
" deunawiad", fal y geilw gwŷr Morganwg eidion blwydd a hanner oed
(alluding to deunaw, /.<., 1S months old).

- Arch. Camb.y Itli Ser., iv. p. 269.


[11.] Nouns of relationship cliffer greatly in the dialects.

Cf. Eanes y Ffydd, 1677: N.-W. nain = S.-W. mam gu; do., Y
Gwyl, vi (1828), p. 207.

R. Morris, Add. MS. 14,945, f . 2476: hendaid in Anglesey =
tad cu in Powys; also gorhendaid and hn hendaid = hendad cu.

L. Morris, Add. MS. 14,923, f. 1336: S.-W. mam gu a thad cu
= N.-W. nain a thaid. ld.,ibid., 15,025, f. 806: in Fowys tad d,
mam dd, ' grandf ather, -motlier'; tad c, mam g, ' great-grand-
father'; etc.; nain, ' great-great-grandmother'; mam wen, ' step-

Y Traetn., iii, p. 12: Vened. taid, naiu = Powys. tad da, mam
dda=S.-W. tad cu, mamgu; Vened. tad yn nghyfraith (mam, mab,
merch yn ngh.) = Powys. and S.-W. chwegrwn (chwegr, daw,
gwaudd); Vened. tad yn nghyfraith (mam, etc, yn ngb,) also =
Powys. tad gwyn (mam wen, etc.) = S.-W. lysdad, llysfam, etc.

I may quote from L. Glyn Cothi's Poems, p. 210: Tir yr
hynaiv, trwy raniad | A rhau o dir yr hen dad; | Tai'r gorhendad,
a'r tad da | Tai'r ewythyr vl Troia. (He uses broder, besides
brodyr; cf. Y tri broder lle gosoder | Yr aur doder ar wyrdedwydd,
p. 42; broder Rhosser, p. 43; Dau vroder, ryw amser, oedd | A
wnaeth Ruvain a'i threvoedd, p. 433, etc.)

Daw, pl. dawon, Davies, Gramin., p. 40; and dofion.

Altrou, ' victricus', aud altruan, 'noverca', Corn. Vocab., are W.
alltraw, ' sponsor', and elltrewen, ' stepmother'; Davies, Dict., has
elldrewyn, from W. Lleyn's Vocab.

Cefnder, etc, are pronounced at Neath: centar, pl. cenderwydd;
cnithtar, pl. cnithterwydd; cyferddar (= cyfyrder); ib.: whr,
whiorydd; brawd, brotyr; tad yn nghyfrith.

In Breton tad coz is ' grandfather', tad cufi, as in Powys, ' great-
grandfather'; tad you, ' great-great-grandfather'. Tad caer, ' father-
in-law' aud ' stepfather' (tadec in Vann.), is in imitation of the
French name; the Welsh tad gwyn shows the same idea. It
is lestad, as in S.-W., in the dialect of Lon. Deufi, gouhez are
W. daw, gwaudd. Tadek, gourtadik exist in Batz (Ernault, Batz
Dialect, p. 34); cf. W. gorhendad.

A list of all possible degrees of relationship, with the Welsh
names, will be found in the Cambrian Jonrnal. It was sent to
Wales by Lewis Morgan and filled up by Williams ab Ithel.
Though Morgan's circular, showing the kind of researches he was
making, is printed before the list, Ab Ithel gave him only the
literary words, and omitted the dialectal forms, different as they are.
Morgan's great work is well known, and so is the German book of


Engels, which is based on it, and the perusal of such works
would, by the way, induce me at least not to use the words of Kuhn's
Zeitschr., xxviii, p. 421, n. 1, if 1 understand them right, with
regard to the Irish old ages in general.

[12.] Common abbreviations, etc, of names are:

Jack, Jacky, Jacko, Siac, Siacci; Sion, Sionyn; Jenny, Jinny,
Jinno, Jinnten, Sin, Sini, Sieni, Sianten; Wil, Bil, Bilo, Bilws,
Bili; Catrin, Kate, Kit, Kitty, Catti, Cadi, Cadan, Cadws, Cadsan,
Cadsen, Kitsen (F Traeth., iii, p. 14); Dai in S.-W.; Palws,
Malws. Mali in parts (Mli, Mlen in Dyfed), ' Mary', Cambrian
Journal, iii, p. 243; Twm, Shn, Dai, Mocyn, Harri, Wl, Nd,
Palws, Sl, Magws (' Margaret'), ib., iv, p. 37. '

1 A passage from this article on the lines of demarcation of the Gwen-
tian dialects may be quoted, as it seems to be based on actual observation,
and contains facts on which others raay supply further information:
"There is a great difference between the dialects of Menevia [= Myn-
wy] and Morganwg. Throughout the middle and eastern districts
the vowel i has almost its full sound in hundreds of words, as shall be
noticed hereafter. Towarda the Saxon border, a certain strangeness
dwells on the faces of the men, somewhat similar to the gloomy appear-
ance that ensues when the sun is hidden by a cloud previous to its setting
in the west. From Ergyng to Talgoed (Caldicot) one meets with heavy,
lanky,andveryignorantmen; and theold people thatarethere,especially
towards Tre'r Esgob [Bishopston, near Newport], speak Welsh, which is
unintelligible to the uui-lingual Cymro. They have so much of the
English accent, and occasionally an old word like bargofi, that they
cause a mixture of grief and astonishment in the bosom of the visitor.
When he proceeds from Crughywel to Cotd y Cymmer, he hears
clearly the accent and pronunciation of the Brecknockian; ar yr nn [in
Glani. dan yr un, sc. awr, ' at the samc time'], hul raig [the infected
forms of gwlad, gwraig] ferch y forwn, etc, present themselves
there very distinctly. When we go from Coed y Cymmer through
Cwmamman to Pont ar Ddulas, we hear thepronunciation of the
Brecknockian, and that of fche boys of Caermarthen. Here the speech
becomes vigorous, and the voice fchin; and yn wironedd fach anwyl i
[pron. yn wirione fach anwl i], thinci fawr, come to light; and, in
returning, achange willbe perceived towards Margam, and a little after
towards Pont Faen [Cowbridge]. Then the body of the country is
reached, and the tone becomes slow and grave, the tongue lisps a
little, and the voice is thick. Abertawy [Swansea], Merthyr, and all
the works, Cardiff and Newport coutain people from every


[13.] The feminine forms of the adjectives, as trom, gwen,
are beginning to disappear from the living language. Davies,
Gramm., p. 54, says that chwyrn, gwymp, gwydn, hyll, syth,
tynn, and clyd were then used in N.-W. as masc. and fem.
forms; but that in Powys (p. 59) the fem. forms of com-
pound adjectives in -lyd, as brycheuled, poethled, were still
used (cf. fford lysseulet, ' a fowery road', Y S. Gr., p. 220).
The fem. forms sech, gwleb are said, in a letter from John
Morgan to Moses Williams, copied in several MSS. (Add.
MS. 14,934, f. 1756), to have then been common in Anglesey.
On the present language see Eowland, Gramm, p. 41; and
Sweet, p. 438.

[14.] In the same way the few isolated comparatives and
superlatives like iau, lled, etc, are being replaced by modern
formations in -ach, -af, as ieuangach. Trechach, given by
Davies (Gramm., p. 63) from Sion Tudur, is said in a note to
the 2nd ed. (1809) to be quite common at that time (p. 81);
and also lletach for lld.

Lled is retained by the Gwentian dialects. Cf. Y Traeth., iii,
p. 14: lled na'r ddaear, N.-W. lletach. The adverb lled, ' partly,
alrnost', takes the position of the N.-W. go. Cf. Hugbes, 1823,
p. 33: lled agos, lled dda; Camb. Journ., iii, p. 252: lled od,
Monm. and East Glam.; lled hynod, West Glam. go hynod,
N.-W.; so lolo Morganwg, Y Cijmmr., iv, p. 105, writes: yn dŷ
ffermwr 1 lled dda.

[15.] Welsh -ach, compared with Breton -oc'h, may

country" (pp. 37-8). These rough notes on sub-dialects, though they
may be known to Welshmen, again give rise to the regret that next to
nothing has been or is being published on the great variety of dialects
in this or, in fact, in any other part of Wales.

i S.-W. fferem N.-W. tyddyn, Hughes, 1823, p. 36. On tyddyn
(tegdyn, MS. A; tgdỳn, Latin Laws, p. 788), and also on tyn, see
Rhŷs, Rev. Celt., vi, p. 49, n. But what are syddyn, eisyddyn,
essyddyn, yssyddyn, given by Davies (Dict.) as Dimetian for Vene-
dotian tyddyn? They oecur in W. Lleyn's Vocab.: eisyddyn =


have assumed a from -/ of the superlative; but the exist-

ence of a similar clistribution of a and o in verbal termina-

tions (see Y Cymmr., ix, p. 73) makes tliis doubtful. Neitber

the existence of -ac'h in the Bas-Cornouaillais (iselac'h,

uhelac'h, brayac'h, etc, given in Eostrenen's Grammar) nor

the occurrence of uchof in old Welsli MSS. furthers the solu-

tion of this cjuestion, as the phonetics of this Breton sub-

dialect, in which the change of o into a may be quite

common, are not known, and uchof does not seem to be any-

thing but a combination of uchaf and ucho, uchod. Huchof

occurs in MS. , pp. 2, 3, 5, 24, 28, 52, etc; uchow, MS. E

(Add. MS. 14,931), ff. 35, 42 (cf. sew, f. la = sef). Uwchel

occurs more rarely. Cf. MS. Caliy. A 13: wuchot, f. 152&,

map wuchelr, f. 177, b; Sal., N. T. j ywchel, ff. 380,

381, ywchter, f. 389; Add. MS. 14,986: ywchel ievstvs,

'. 11, ywchel ddydd, f. 12.

[16.] -Ied, -iach are used with some adjectives for -ed,

-ach; with rhaid, santaidd, etc, nearly always.

Cf. Add. MS. 14,869, f. 131a: o welet vy lle ar llet eithyaf;
Ll. Gio. fA., p. 75: yn gyiiYonedigeidiet; Y S. Gr.: cynsanteidiet,
reidyach, dir(i)eidyach (kyn reittet, ih.); Y Dnjch Chr.: gwelwch
reitied yw mefyrio; Davies, Gramm.: rbaid, rheittiaeh, rheitied;
Cab. few. T., p. 30: er sauteddied rwyt ti; S. C, i, p. 332:
mor brafied, etc.

[17.] -Ed, the existing opinions on which will be found
in Rhŷs, Lect. 2 , p. 231, and in Toyail Troi (LL.), pp. viii-ix
(Lndcx, s. v. ' dubithir'), is by no means a " propria canibrica
ilinlccti forma" (Zeuss, G. C. 2 , p. 931), but is also common in
the Breton dialect of Vannes. Cf. the Gramm. of J. Guil-
lme (1797-1857), recteur de Ivergrist, a native of Malguuac,
translator of the extracts of the Annales de la Foi published
in the dialect of Vannes, and author ol' />< vr eul labourer, 1849,
etc (mentioned in Z. 2 , p. xliv): na brasset- hon puissance,
na biannet ha er jou-ce, p. 122: na carret-, p. 125. In
Rev. CclL, iii, -ct is given rom Sarzeau, and (iv, p. 145) the


-acl of Lanrodec is taken to be a combination of -et ancl

the superl. -a, Cf. Gourhemneu clou ha r en ilis (Vannes,

1879): na brasset- dalledigueah certaen tud ! p. 170. Some

examples are also giyen by Loth, Mm. Soc. Ling., vol. v.

This comparative in -ed is used in comparisons with cyn

(cy), can, in the same sense as mr with the positive (Old W.

mortru,gl. 'eheu', morliaus, gl. f quam inultos' 1 ), and mor witli

-ecl occurs also in the later language, being commonly attri-

buted to the Southern dialects. Cf. Davies, Gramm,, p. 65

Demet.mor hardded = Vened. mor hardd; Dosp. Ecl, p. 259

Vened. and Gwent. mor ln = Dimet. mor laned. Cf. S. G.

mor cheped (cheap), i, p. 373; mor gynted, iii, p. 226. Y

Fcltcn: morbellad, Dec. 23, 1870; YGweithiwr: mor belled

ag w i yn deall, No. 1; this is, however, from Gwentian


n the use of -ed, cf. also B. of Herg.: hyhelaethet, col. 570 2;
yn gyainlet and yn gynamlet, col. 606; py gy bellet odyma y y
cruc a dywedy di, col. 681; ac a dywedaf itt py gy bellet y6, ib. 3;
Ll. Gw. Rh.: y gystal na y ganfuanet, p. 127; yr hwnn .... bot
yn gynn amlet y wyrda ac yn wychet, p. 2; rac mor ieuanc oed
a gwannet y hannyan, p. 16; with er, yr: MS. Tit. D 22, f. 12a
(Y Cymmr., iv, p. 120): yr tayred vo yr heul; Y Drijch Chr.:
er drycced a fo'r dynion, f. 19; Cab. f'eic. T.: er gwaethet
oedden nhwthe, p. 60; er santeiddied rwyt ti, p. 30. Daf. ab Gw.,
p. 103: Doe ddifiau, cyn dechrau dydd, | Lawned fum o lawenydd !

[18.] The following notes contain some additions to Zeuss'

remarks 3 G. G.' 2 , pp. 616-20, on Adverbs of place, time, etc.

Diuinid, ib., p. 616 (' sursum'), Lib. Land ,frcq. Ar i fyny, tuag
i fyny, Sp., Eng.- W. Dict., 3 s. v. ' upwards'. Y uynyd is the regular
form used in the R. B. Mab. (cols. 560, 566, 567, 4 etc), and in
otherS.-AV. MSS.; e.cj., MS. U, p. 355; Harl. 4353, f. 35?;, etc.
It is later limited to S.-W. dialects, cf. E. Lhuyd, s.v. ' supra' 5;
Gwallt. Mech., Worhs, ii, p. 214; i fynydd, Homil., 1606, i, p. 105;

1 Ovid Glosses, 39a; Stokes, Cambrica, p. 235.

2 R. B. Mab., p. 160, 1. 26. 3 Ib., pp. 222, 1. 1; 223, 11. 29, 30.

4 lb., pp. 149, 1. 27; 155, 1. 21; 156, 1. 25.

5 Arci. Brit., p. 159, col. 1: ll Dimet. i vnydh".


i fynydd [Monmsh.], Sal., N T, Matt. xvii, 27; Acts xiii, 37;
James i, 17; Rev. iv, 1; etc.

Y Traeth., iii, p. 14: N.-W. i fyny, uchod = S.-W. fry, fry ar y
lan (glan, in its S.-W. meaning of ' hill'); Rhỳs, Loanwords, s. v.
' altus' 1: N.-W. i fyny i'r allt = S.-W. i'r rhiw, ' up hill'.

Myny has not been explained, for the loss of -dd in this position
is, if we except the unexplained eistedd, eiste (perhaps for eisteu,
like bore, etc), quite isolated, save in Dimetian dialects (gwirioue,
dy, towi, etc). I suppose that the last syllable has been assimi-
lated to bry, fry. Mynydd, often written monyth in Salesbury's
N. T. (ff. 6i, 7a, etc), a pronunciation given as that of St. David's,
Pembrokeshire, is mwni in the present Dimetian dialect; see on
this point Dosp. Ed., p. 257, and S. Gomer, i (1814), 19, where
idiomatic expressions like myn'd i'r mwni 'r gwartheg (for cae),
i'r mwni i aredig (for maes), i'r mwni i hau, i fedi are given,
and where it is said that every trofa cyffredin (' common')
would be called ' mynydd' in Glamorganshire.

Dirguairet, Z. 2 , p. 616 ('deorsum'), Lib. Land.,freq.; Y Drych
Chr., f. 75a: i fyny ag i wered; YTraeth., iii, p. 14: N.-W. i lawr,
isod = S.-W. obry, i waered; L. Morris, Add. MS. 14,944, f. 66a:
dobry, common in Cardiganshire for dy obry. Y Traeth., I. c.:
N.-W. i lawr, ymaith = S.-W. i bant; i bant ag e; L. Morris,
Add. MS. 14,923: S.-W. i bant = N.-W. i ffwrdd, i ffordd, f. 133a;
S.-W. godir, ' a hollow place between valleys' = N.-W. pant,f.l33.

According to S. Gomer, 1814, l. c, godir is the S.-W. word for
pant, ' valley'; and pant has so lost its real meaning thatone would
say: ' bod gwr wedi myned bant i'r mynydd, ac i bant i'r bryn.'
Cf. also Hughes, 1823, p. 37: S.-W. i bant = N.-W. y ffordd (i
ffwrdd is S.-W., accordingto Rhŷs, Lcct?, p. 114); Y Geninen, iii,
p. 19: (S. W.) i'w gl i bant = ei symud i ffwrdd.

Y maes, Z. 2 , p. 616 (' foras'), Mb.; L. Morris, Add. MS.
14,944: S.-W. myn'd i faes=myned ymaith, ' to go along'; id.
14,923, f. 133/;: S.-W. troi maes, 'to turn (to windward) out'
= N.-W. troi allan, hwylio allan; Y Tracth., iii, p. 12: S.-W. i
maes, i faes = N.-W. allan; L. Morris, /. c, f. 133: S.-W. i macs
o law, immediately=N.-W. tocc, yu gwit. Cf. N.-W. toc a da,
' presently and in good time', RoAvland, Gramm. 4 , p. 114, and Yr
Arw., July 17, 1856: A toc dyma, hi yn mund i dywallt cypanaid
de i'rhogun; a dyno fo yn gofun toc; etc.

N.-W. allan o law =S.-W. maea o law, ' presently', Rowland, /. c.;
for i maes and i faes, cf. N.-W. i mi, S.-W. i fi, Y Cymmr., viii,
p. 139.

i Arch. Camb., Ith Ser., iv, p. 262.


[19.] Ymywn, Z. 2 ,p. 616 ('in meclium et in medio, intra, in').
Neither the loss of dd in niewn, mywn (Ir. medhon), nor the
Venedotian and Powysian forms meawn,miawn,nor the accent
which causes S.-W. mwn for mewn, are sufficiently explained.
Mwn probably arose from mewn being pronounced as an un-
stressed proclitic before nouns. For examples see my Notes
on W. Oonsonants (Rev. Celt., ix, pp. 64-76, etc).

[20.] Yr meitin, Z. 2 , p. 616 (' paulisper, paulo aute'). Mah.:
ysgwers, ystalym (' dudum, iam dudum'); Hughes, 1823, p. 38:
N.-W. hawg, yr hawg = S.-W. ys smeityn; but cf. as myitin,
Yr Arw., Jan. 20, 1859; es meityn, Cab.f'ew. T, p. 65.

Y Traeth., iii, p. 13: N.-W. yr hawg = S.W. enyd o amser;
N.-W. yr hawg iawn = am hir amser; Cambr. Journ., iii, p. 252:
Monm. and East Glam. smityn, West Glam. smeityn = N.-W.
hawg, yr hawg.

N.-W. (L. Morris, Add. MS. 14,923, f. 133) ers talwm = S.-W.
ers llawer dydd; Y Cymmr., iii, p. 84: colloqu. ystalwm, 'stalwm;
Ll. y Resol.: N.-W. ystalm = S.-W. ysdyddie. Cf. Can. y C, p
378: er ys dyddie.

Er ys is written ar's, a's in some N-W. texts; cf. Yr Arw.:
ars llawar o amsar, Dec. 11, 1856; as talwm, July 17, 1856; May
28, 1857, etc, like ario'd (see 22).

In S.-W. os, perhaps not for oes, but f or o'r ys, o'r's; cf. S. C.:
ys lawer dy, i, p. 232; 'slawer dy, iii, p. 324, ib.; os dyddie,
also os ticyn yn ol, os cettyn bach. L. Morris, Add. MSS.:
cettyn, a small matter, Cardigansh., 14,944, f. 46; S.-W. cettyn,
a good deal = N.-W. twysgen, 14,923, f. 134a; cf. yn well o
getyn he'yd 'na, P. C, 28 (Ebbw Vale).

Richards, Dict., however, has: twysgen, ' a small part'; in S. W.
twysged, 'a good part, a great deal'. Hughes, 1823, p. xi: S.-W.
cettyn = N.-W. darn; Y Traeth., iii, p. 13: N.-W. tipyn o ffordd =
S.-W. cetyn o ffordd; N.-W. cryn dipyn o ff . = S.-W. cetyn diogel
o ff. Cf. Hughes, 1823, p. 38: N.-W. Gryn = S.-W. iawn; gryn
of n = ofn mawr; Yr Arw., Feb. 12, 1857: cryn lawar; Add.
MS. 14,923, f. 134a: S.-W. yn ddiogel, 'for certain'= N.-W. siwr,
yn siwr, yn ddiammeu.

[21.] Drachefen, drachefn, Z. 2 , p. 616 ('trans tergum, retro').
Cf. dramkevyn, Y S. Gr., p. 184; drach dy gevyn, p. 275;
drach eu k., pp. 283, 301; drache cheuyn, B. of Herg., col. 866;
dracheukeiyn, Add. MS. 19,709, f. 30; darchefyn, Jes. ColL


MS. 141, f. 14&; yno ydd ymchoelodd ef trach i gefyn, Sal.,
JV. T., f. 266; drachefn, drachgefn, Davies, Dict.; drychgefn,
dychgefn in modern dialects, since -gefn bears the accent.

Trach occurs in trach lavnawr, B. ofAn., God., St. 79 (Sk.,
ii, p. 86); oes tragoes, B. of Herg., Sk., ii, p. 230. Tra chaer
wydyr, B. of Tal., jSTo. xxx (Sk., ii, p. 182).

Cefn gefn, ' back to back' (Anglesey), Add. MS. 15,025,
f. 80. Cf. daldal, Mab., ii, p. 54 (R B. Mab., p. 285, 1. 12);
benndraphenn, Y S. Gr., p. 344; ben dra mwnwgl. Cf. Add.
MS. 15,027, f. 816: fe aeth ein gwlad ni (fal y dywaid trigo-
lion Deheubarth) yn bendramwnwgl (Lettcr to Owen Jones);
I have met also with a verb formed from this expression.

[22.] Eirmoet, eiryoet, Z. 2 , p. 616 (' unquam'). Erioed, Sp.,
Bict?, yrioed, irioed, Sal., N. T; irioed, Add. MS. 31,055,
f. 200a (Dr. Thos. Williams); yrioed, ib. 31,057, f. 169.a; S. O.:
ario'd, ii, p. 262; iii, pp. 384, 447; N.-W., yrioud, rioud, Yr
Arw., etc. Cf. er and yr (duw), etc, the phonetics of which
are also obscure to me as regards their dialectal distribution.

Ysgwers, Z. 2 , p. 616('dudum, iam dudum'); Richards, Dict.:
gwers, 'a while'; ym pengwers, 'a whileafter', S.-W.; see YCymmr. f
viii, p. 152: S.-W. yn awr, 'nawr = N.-W. rŵan; L. Morris, Add.
MS. 14,923, f. 133/>: S.-W. ynawr ag eilwaith = N.-W. ynŷan ag
yn y man, 'now and then'; YTraeth., iii, p. 12: S.-W. awr ac enyd
= N.-W. byth a hefyd.

S.-W. awr ac orig (e.g , Can. y C, p. 461, bob amser' in the
margin) = N.-W. o hyd, trwy gydol yr amser (ib.). N.-W. yn
uuion deg, ' imniediately' (colloqu.), Rowland, Gramin, p. 114.

Ilnghes, 1823, p. 35: S.-W. bron, o'r bron = N.-W. yn rhestr

(sic). Braidd, 'just, hardly, scarcely; nearly, almost', Sp., Dict. 3; L.

Morris, Add. MS. 14,944, f. 386: braidd, ' hardly, scarcely'; braidd

na, ' prope', " corrupt in N. and S. W. braint"; a'so ib., . 40a: 1V a

'general corruption''; so: braint na luddasai 'r naill y llall.

Brain really occurs; cf. S. Gomer, 1851, p. 99: brain y gaffo
i, (ilamorgansh.; P. C, Jan. 22, 1859: brain o beth yw segura
fel hyn, Ebbw Vale. 1 suppose that ' braint' is merely an
1 1 yiuologising orthograihy, and that brain represents braidd-na.

Anglesey, Add. MS. 14,944, f. 576: ar y cyngyd, ' just a doing';
Can. y C, pp. 379, 400, 569: 'n immwngc, explained in the margin


by ' disymwth'; cf. yngo, 'hard by' (Dosp. Ed., 901), e.g., Pettwn
hebddo, yngo angerdd, Daf. ab Gw., p. 255; Yngo o Heiwordd
hyd yn Nghynvig, L. Glyn Cothi, p. 36; wnc, ' hard by', Dosp. Ed.,
I. c.; echwng; rhwng, see Y Cymmr., viii, p. 129.

Hayach, haeach, haeachen, passim haechen, fere, Davies, Dict.,
haiach, haiachen, ' instantaneously; almost, most', Sp., Dict. 3; hay-
chen, Davies, Gramm., p. 147; hayachen, Mab., ii, p. 247 (R. B.
Mb., p. 142); heb wybod haychen, beth yw . . . . , Y Drych Chr.,
f. (3); ar haychen (marg. agos) boddi, Can. y C, p. 340; haychen
(marg. agos, oddieithr, ychydig), ib., p. 350; haychen (marg. ym
mron, agos), ib., p. 538.

Ychydig, now commonly chydig, is often written bychydic in
Middle-W.; on this word (Ychan for Fychan is also often met
with) see my Notes on W. Cons. (Rev. Cet., ix, p. 76, etc).

Rhagor (subst.) in N.-W. is gwahaniaeth (Ll. y Resol.), ' differ-
ence'; in parts of S. W. rhagor (adv., ' more') is [the same as]
ychwaneg (chwaneg), Richards, Dict., Hughes, 1823, p. 36, etc.

[23.] To emphasise an assertion several adjectives like
ofnadwy, cynddeiriog, anghomon ('uncommon'), ffiaidd, etc,
with yn are used in vulgar language in different dialects, and
sometimes interesting phonetic alterations have been made.

Sweet, p. 431, gives novntsan, which he thinhs to be for yn
ofnadwy faswn. Ffaswn (Eng. ' fashion'), when unstressed loses its
first syllable; cf . y mae bud na welist ti rioud siwn beth yma, Yr
Arw., Dec. 11, 1856; welis i rot siwn beth, ib., June 20, 1859;
rptswn, ' ever the like, ever' = riyd faswn, Sweet. I do not,
however, think that it is contained in ofnadsan, although 1 cannot
explain the latter, or do more than give a number of other altera-
tions of ofnadwy.

Cf. Y Traeth., iii, p. 12: ofnadwy, ofnaswy, in Ardudwy (Meri-
oneth) afnadsan; Yr Arw.: ofnadsan, ct. 30, 1856; ofnatsan,
March 3, 1859; peth rhyfadd ofnedsan, Feb. 12, 1857 (sic); mi
gafodd le da afnadsan, yn rhyfedd afnadsan gini, May 7, 1857;
S. Cymru: ofnatsan, iii, pp. 103, 186; Y Gweth.; ofnaclsen, 1858,
No. 1 (Aberdare); Y Bed.: wyt ti 'n depyg afnatyw i dy dacl,
viii, p. 106 (Monmouthsh.); yn ofnatyw, p. 174.

In Neath afnadw; afnaswy, YrArw., Dec. 11, 1856. Yn ofnad-
widd (Yr Ams., Jan. 14, 1847, in a S.-W. letter, on which see
Y Cymmr., ix, p. 118) is probably more than a varying ortho-
graphy (as towi might be said for tywydd, the final -dd not being
pronounced), for a native of Anglesey was famihar with it too.


The change of o and o is frequent in the dialects. See on this
poit my Beitr., p. 52; and cf. eidrol ( = eidral), Lhuyd, Arch.
Brit., p. 279, s.v. ' Ivie'; molwan, pl. molwod in Anglesey (malwan
in Carn. = malwen, ' snail'), afol in Anglesey (afal in Carn., like
gofol, dafod, Beitr., 57); andros in Arvon and Angl., Rhŷs, Ilibb.
Lect., p. 200, n. 1; myolchan in Neath (= mwyalchen), anglodd,
ib. (= angladd), etc. But in afnadwy I really think a to be
older, since *afan would explain the S.-W. ofan. besides ofon
(Beitr., 63).

Shoe and shaw are used in the same sense. Cf. ac yr w'i yn eu
licio yn she, S. C., i, p. 411; yr ydwi yn ffond shoe o ydrach ....
ib., ii, p. 487; a'u bod wedi gueyd shoe o ddrwg, p. 186, etc. (Merio-
nethsh. dialect); shoy o Sbarthwr (= dosbarthwr) i chi, dduliwn
(= feddyliwn) i, Yr Ams., Nov. 16, 1848.

P. C, 29: ac y mae yn shaw o gwiddyl (= cywlydd, like giddyl
= gilydd) iddi nhw; y ma' yma beth shaw o de a theisan; shaw o
ddioni (= daioni, daoni), Ebbw Vale; Yr Ams., Jan. 14, 1847: 'n
show, S.W.; Y Gwron Cymr., May 6, 1852: ma rhwy shew o beth.

Shaw, a ' great deal' (afnadw, ofan in Neath). If shoe is not the
last syllable of ofnaswy, I cannot explain it.

In S. Gomer, i (1814), 19, are quoted from the Dimetian
dialects: caru merch yn ofuadwy, yu embydus, edrych yu ofnadwy
(= yn graff) ar eu gilydd, merched yn ln (or bert) ofnadwy,
embydus (= yn brydweddol, yu landeg iawn). Twenty years
before (that is, in the last century) ffamws was often used in these

In Glam., merch ln fudr, benyw lan fudr. With budr (' dirty,
nasty, filthy, foul, vile', Sp., Dict?) cf. Glam. budyr = N.-W.
cethin, Ilughes, 1823, p. 34; N.-W. budrog = puttain (S.-W.),
Hanes y Ffydd, 1677; S.-W. brwnt = N.-W. budr; S.-W. soga =
N.-W. dynes fudr, Y Gicyl, vi (1828), p. 207.

Cf. also Cambr. Journ., iii, p. 246: Dimet. mai 'n dewi ŵer iawn,
ymbeidis, embydus (in S. C. also printed ombeidus) = West Glam.
niai 'n dywydd r (Monm. and E. Glam. gr) iawn (or fine unco-
mon); Dimet. merch lu iawn (or odiaeth) yw hi = West Glam.
m. ln iawu, sometimes (so Monm. and E. Glam.) ln fudyr.

Ebbw Vale, P. C, Feb. 4, 1860: bydd yn dda budr, fe fiiodd
slawer dydd yn dotal budr; ib., anghmon and yn greulon are used.

Iolo Morganwg remarks, in Cyfrinach Beirdd Yn. P>\, 1829,
p. 238: y niae .... yn ymhoffi 'n fawr (nou >/ ffaidd, fal y
dywedant ym Meirion; ynfudr yw gair Morganwg). Y Traeth.,
iii, p. 12: dynes ln arw (garw). Cf. wedi gacl blesser garw
iawn, )V Ams., Dec. 3, 1846; bod arni eisio y wlanen garw iawn,
ini gcfis helunt garw iawn hefyd hefo dyn, Yr Anc, Dec. 11, 1856;


Y Traeth., I. c.: dyrna beth clws (for tlws) ofnadwy, gwych aflawen,
da gynddeiriog, cryf anafus; cf. also yn rywinol iawn, Yr Ams.,
Sept. 10, 1849 (gerwinol); yn ryswydus, Nov. 29, 1849 (= arswy-
dus), etc.

[24.] ISTauiyn, Z. 2 , p. 620 (' tantum'), is the modern form,
and also that most frequent in Mid.-W., but atnyn,
yn amyn, and namwyn occur also in older texts. Cf.
MS. A: anien, p. 46 (4 times); amn, p. 125; also in MS. B
(Tit. D 2), f. 37: amn. Namin, A, pp. 57, 58, 59, etc.;
in these pages i fory, oftener writtene, is especially frequent;
cf. gustil and guestel, gustel, kamrit, idin, aegilid. Namen,
pp. 3, 46, etc. Naman, p. 58 (thrice), p. 66. Cf. on these
same pages kafreis, llana, kanas, pp. 57, 59, etc.; see my
Beitr., 33. Kannas (= canys) occurs also four times in
a late 15th century fragment in Hgt. MS. 57 (the frst three
pages of an otherwise unlmown Welsh version of, appa-
rently, Perceval le Gallois), in which also occur yneidiev and

Namuin, p. 58; namuyn, p. 59, etc.; namun, p. 58. The
last (cf. racu, ib. (= raccw), gustlaf) represents namwn from
namwyn. On p. 58 alone there thus occur 1 namin, 3 naman.
1 namuin (1 namun). Cf. B. ofCarm.: namuin, Nos. v, ^"yii 1;
B. o/Herg.: namwyn, Sk., ii, p. 249; namyn, col. 1186 (Poem
of Gynnuard Brecheinyac); Ll. Gw. Bh.: namwyn, p. 135
(Bown o H.); MS. CT.B5: namwyn, ff. 178, 191, 2166;
nammwyn, f. 175; Add. MS. 14,869: drudyon a veirtyon
(avawl neb dragon)namwyn dreic ae dirper,f. 80 (Cynddelw);
namwyn, f. 109; ny chaffad gwrthep namwyn gwyrtheu, f.
113 (Gwynnuart Brycheinyawc); Add. MS. 19,709: amynn,
f. 17; MS. S: ynamyn, f. 90.

Sal., iV. T.: amyn, ff. 52a, 54, 626, 716, 77, etc.; n'amyn,
ff. 526, 1276; yn amyn, f. 305; y namyn, f. 1116; namin,
f. 336; namyn often. Iolo MSS., p. 253: namn ei eni
(Chwedlau V doethion, No. 32). Pughe gives naniwyn, namyn,
nam, amyn; named fi ( cf. Vann. Bret. nemedouf, etc), etc.
1 Ff. llfl and 25a; Sk., ii, pp. 8, 19 (respectively).


The explanation of these words presents great diticulties,
and I can only proffer a few guesses. The existence of the
Corn. lemmyn (not from *lle-amyn, but from *nemmyn by
means of dissimilation) and namnag, Bret. nemet, nemed
(Vann. nemeit, meit), with suffixed pronouns, like nem-of in
Welsh (see YCymmr., viii, p. 127), make it at least probable
that nam-yn, not aniyn (*am-hyn?), is the older form, though
the contrary view is not untenable, as mae, explained by
ym-ae, has its counterpart in Breton too. The Ir. namm
strikes us at frst, but its regular connection with acht, ' but',
has suggested the explanation acht na-m-b, containing the
relative pronoun; and the Ir. amn, amne, ' so', might ecjually
well be taken into consideration.

Nam, -au (Sp., Did. z ) is ' a mark, maim, fault, exception',
nainu, ' to blemish; to except'. Viewed from this side namyn
might be *nam-hyn, like noson from *nos-hon. At any rate,
*am or *nam remains, and I will now consider the relation
of namwyn to namyn. If I am right in comparing eiswys
and eisys (see below, 26), namyn may be the outcome of
*namwn from namwyn, the changes being due to the shifting
of the accent; for the verbal ending -wys, -ws has never
yielded *-ys. However, the frequent position of these ad-
verbs as proclitics or enclitics may have produced, under
circumstances not exactly known, such doublets as eiswys,
eisys. 1 If I am wrong in this, and if aniwyn has also nothing
to do with Ir. amn (J and wy from *), then I can only adduce
etwaeth, oddynaeth (ynaeth, Rgt. MS. 202 (bis), f. 22b 2; Y
Cymmr., vii, p. 125), in which *-ac-to- has been found, and
explain namwyn from *nam-wg-n (like dwyn, amwyn). I
really think this last the most probable of all these suggestions,

1 Although they are not connected with the prcsent cjucstion, I will
mention hcre ellai besidcs fallai (for ef allai, 'perhaps'; in Anglcsey
hwrach); acha bora (= ar ucha bore), Glam.

Inrheg Urien, 11. 7, 11 = Sk., ii, p. 292, 11.:, 7.


and it would not exclude the comiection of nam-yn with

[25.] Nachaf, nychaf is one of the words in which a pre-
fixed preposition ceased to he felt to be a separate word. Cf.
ynachaf, nachaf; ynychaf, Ll. Gw. Rh., pp. 21, 25, 28; y
nchaf, MS. Gl.
B 5, f. 10; Sal., N. T.: nycha, f. 3;
nacha, f. 1716; nachaf, 1776.

Davies, Dict., has an obsolete ycha, ' en, ecce'. Nachaf is
yn with the superlative of ach. Cf. Z. 2 , p. 694: ach y law;
ach: ych like am: ym-; B. of An. 1: ech e dir ae dreucl. I
look with more confidence on " ech corrupte pro edrych, Car-
narvonsh. and Montgomerysh." (L. Morris, Add. MS. 14,909,
ff. 52., 69 (thrice) ), since I have found it frequently used in
Hanesion o'r Hcn Oesoedd, 1872 (Carnarvonshire dialect), e.g.:
mi af yn awr i ech am dano, p. 60; dwydwch y doi i e'ch am
dano, p. 66; pan eis i'r tir ac e'ch i fynu, p. 150. Davies' ycha
might perhaps belong to it, unless this is abstracted from yn
ycha, and has no separate existence at all.

[26.] Eisoes, eiswys, eisys, adv., ' lihewise; already', Sp.,
Dict? Cf.eissoes,eissyoes,the common Middle-W. form; eisys,
in MS. Tit.B 22,f.l0, is given by Powel(F Gymmr.,iv, p.107)
amongst the Dimetian peculiarities of this MS. Perhaps
" South-Welsh" in general might be said, for cf. Ll. Gw, Rh.
eisswys, pp. 188, 204, etc.; MS. Gl. B 5 (Gwent. dial.)
eissiws, f. 326; eissws, f.41; eisws, f. 50; eisswis, f. 78
Barddas, i: eiswys, p. 78, etc.; Can. y G., p. 430: eiswys (in
rhyme with prynwys, pron. prynws?); Sal., N. T.: eisus
{marg. esioes), Gwel.L, f. 376; eisius, ff. 786, 309; eusus, f.
3306 (R. D.); E. Llwyd, Arch. Brit., Bref.: etSS, eysys. 2
I cannot explain these words. Arlloesi is in S.-W. allwys;
dioer, diwyr are unexplained (see my Beitr., 106). On
the possible relation between eiswys and eisys, see 24.

[27.] Ysywaeth, adv., ' more the pity/ Sp., Dict? Cf. Corn.

1 God., Sanza xiii; Sk., ii, p. 66. 2 First and Iast pp.



syweth, soweth, Breton siouaz (in Vann. siouac'h). As
in Corn., ysowaeth, osowaeth are frequent in Welsh. So
yssowaith, Add. MS. 14,986, f. 42 (16th cent.); ossoweth,
ib. 14,973, f. 107 (Araith Ieuan Brydydd Hir}; osowaith,
osywaith, Hope, Pocms (1765); ysowaeth, Can. y C, p. 13;
etto 'soweth, p. 325 (y fo gwaeth, p. 393).

L. Morris (Add. MS. 14,944, ff. 985, 148&) calls it a N.-W.
word; so Davies, Dict, 1: " Venedot. ysywaeth, ' quod magis
miserandum'; Demet, gweitheroedd vel gwaetheroedd." Cf.
gaethiroed du heb 6j, B. of Hcrg., col. 411 (Yst, cle Car.
Mag. 2; and, in the same passage, gwaethiroed duw. lieb
wy, ampeu Charlijmaen, Ll. Gw. Ph., p. 51); a gwaethiroed
nas lladawd, ib., p. 128; gwaethirodd, Stowe MS. 672, f. 556.

A combination of yssywaeth and gwaethiroedd is yssy-
waethiroed, Bown o H, Ll. Gw. Ph,, p. 146; Aeth Herast,
yswaetheroedd ! | Yn drist, L. Gl. Cothi, Poems, p. 38, v. 15,
where the editors note: " yswaetheroedd = yswaetherwydd;
ysywaeth, Alas ! " Cf. gwaetherwydd, ' alas,' Sp., Dict. 3
These last forms recall the problem of eisoes and eiswys.

On agatfydd, agatoedd, etc, see Y Cijnmr., ix, p. 98.

[28.] The preposition oc, o (see Y Cymmr., viii, pp. 135-9),
also ac, a, is interesting on various accounts. On its com-
position with a pronominal element (*son-?) see /. c. Oc is no
longer used, and it is said to have been (iwentian; it is,
however, frequent in all earlier Middle W. MSS.

Cf. oc eu kereynt, MS. A, p. 41; B, ff. 16a, 37; B. of Carm.,
i. 3lb (Sk., ii, p. 27): Megittor oc ev guir. v. hir alanas; Hgt.
MS. 202: oc eu herwỳd (bis), f. 22b; oc eu korff, f. 26 3; B. of
Herg.: oc eu hystoryaeu 6y, col. 229; oc eu hol, col. 202; oc ach
gweithretoed, col. G20; Add. MS. 19,709, ff. 9&, 14, 18/): 6ynt a
wnaetliant aerua diruar oc eu gelynyon, f. 26; ny orffoyssys
Gw. oc eu liymlit hyt pan vei oc eu golut 6y ykyfoethogei ynteu
y teulu, f. 66b; LI. Gw. Rh.: at eu diwreido oc eu chwant. ac eu

1 s. v. 'Gweitheroedd'. 2 Cy mmro dorion edition, p. 30.

3 Y Cymmr., vii, pp. 125 (cf. Sk., ii, p. 292) and 135.


dryc dynyaeth, p. 282; Y S. Gr., pp, 61, 127; MS. Cl. B 5: oc eu parth wynteu, f. 234a, col. 1, etc.

On *oc in rhoc, rhac, see Y Cymmr., viii, pp. 127-8; the Breton raok should be considered too. Cf. raok, Le Gonidec; Trc. en he raoc ha var he lerc'h = Vann. en h raug hag h goud (in the Trc. and Vann. translations of Introd. ad
vitam devotam
); e'rauc, l'A* * *, Dict., s. v. ( devant', e'm-rauc, e'm oun-rauc, etc.); araug hag ardran, Rev. Celt., vii, p. 330 (Vann.); arg, (Bas Vann.). A is frequent in Middle-W. MSS. in expressions like cam a beth, da a was, etc.

Cf. B. of Carm.: maur a teith deuthan, f. lb, Sk., ii, p. 231;

B. of Herg.: truan a chwedyl a dywedyd, Sk., ii, p. 231;

Ll. Gw. Rh.: da a was; mawr a beth, pp. 129, 141, 165; aghywir a beth; glew a beth, p. 125; dewr a was, pp. 125-6; praff a beth, p. 136;
truan a beth, p. 156; ys drwc a chwedyl, p. 165.

For the modern language, I find given in Rowland,
Exerc, p. 143, as S.-W. druan (druan g ef, druan chwi) for druan o'r dyn, druan o hono; and, as used in colloquial language, druan oedd y dyn, druan oeddych chwi, druan oeddynt hwy. Sp. Dict. 3 . s. v. ys, has: ys truan o ddyn wyf fi,
'wretched man that I am.' (Cf. Davies, Dict.: Er asseveratio, Demet. pro Venedot. ys; but perhaps this obseiwation is only abstracted from his "N.-W. ysywaeth, S.-W. gwaetheroedd," see 27.) Druan oedd y dyn looks like a faulty orthography
for druan odd y dyn = druan o'r dyn; cf. the S.-W. use of odd for o before the article, Y Gymmr., viii, p. 146. I cannot, however, decide this question. Druan ydy nhw! Yr Arw., Aug. 20, 1857, points, of course, in the other direction.

[29.] Ac, a, occurs frequently, and in the modern language regularly, before the relative pronouns a and y, with or without the article, whilst oc, o is the Middle W. form of the preposition in this combination. There occur oc a, ac a, or a (o'r), ar a (a'r), and, later on, also ag alone, infecting the initial consonant of the verb it governs.


Cf. MS. Z, p. 275: o pob iar ac a uo ynny ty = or a uo, MS. ./,
= oc a uo, MSS. M, 0, Q, 7 1 , = a uo, MSS. J, S; L, p. 250:
ygkyueir pop kymh6t or y kerda6 drosta = oc k. d., MSS.
Q, T,
= y k. d., MS. S, = a k. d., MSS. J, 0?; L, p. 189: o bop
carcharar oc y diotto heyrn y arnna = or y d., MSS. J, 31, P, =
y d., MS Q; L, p. 191: ympob ty y del = oc y del, MS. J, = or j
del, MS. 31; MS. Q, p. 562: yspeilet ef oc a vo yindana o dillat.

Ll. Gw. Rh.: yny diffrwythont oc an gwelo yn ymlad, p. 98.

Wms., Hgt. MSS., : pawb oc ae darlleo, p. 297.

Y S. Gr.: na dim oc ellit y drossi ar enryded (oc a).

B. of Herg.: ba6b or ae gwelei, col. 613; or a uacker.

L. Gw. Rh.: ar nyt arbetto idaw ehun, p. 2; or a vei reit, p. 3;
or a gaffat, p. 6; drwy arogleu ac eu harogleuei, p. 7 (' for those who
would smell them'); ac a uynnwys . . . . ac ar nys mynnawd, p. 22;
ac ar nyt ymchwelws a las, p. 22; ac ar ny las . . . . p. 26; pawb or
aoed, p. 30, etc.

Add. MS. 19,709: o bob keluydyt or y gellit, f. 9b; ym pop lle
or y bei reit; ar ny ladadoed onadunt, f. 15.

MS. Cl. B 5: or a hanoedỳnt o, f. \ob; ar nỳ las onadunt, f. hb.

Sal., N. T.: bop peth ar y wnaethoeddoedd (sic), f. 208a; o
pob peth ar y weloedd ef, f. 373a. On this use of y cf. Y Cymmr.,
viii, p. 150.

Marchog Crwydrad (17th cent.): pob dillad ag a archei .... ei
gwneuthur, p. 3 (Pt. i, ch. 3); os pob peth ag a gassao y naill fo cr
y llall, p. 2 (Ch. 2), etc.; os is used here as in Add. MS. 14,921.

Can. y C: pob fflineb ag a wnaetho i, p. 63; nid oes uu dŷn
ag 'aner, p. 332; Hope, Poems } 1765: i bob peth ag sydd wrthnebus,
p. 83, etc.

[30.] Like oc: ac (cf. also YCymmr., viii, p. 117), os, ot,or
(' wlien'), which are o, from oc, with pronoininal elements
affixed, occur, but rarely, as a, as, at, ar; these latter forms
are hardly common in any text but Salesbury's N. T.

Cf. a bydd y tuy yn teilwng, f. 15a; a 's byddwch, f. 8a; a 's
dugy dy rodd i'r alLor, f. 7a; ad wyf vine yn ei gwneuthur = os
ydwyf yn eu gwneuthur, ed. 1873. Ani, anid: any bydd, f. 7a;
anid, f. 5a; any darllenasoch, f. 18a (also add ithr, f. 260&, etc.)

In some notes on the orthography followed in Llẁer Gw. Gyffr.,
1586, as reprinted thence in Llyfr. y C, p. 34, " a ' if or whether'
for o", " as for a ys or os'\ are mcntioned.

In the spoken language ys before consonants, 's before vowels,
also ynd for ond, etc, are used. Cf. Cab.few. T.: ys dechreuith hi
son am y beibl, p. 108; 's ydi o'n fyw a tase bosib i mi gael i


ddrecsiwn o, p. 80; Caledfryn, Gramm. 2 , p. 114: ouid oes, pro-
nounced 'un does; Yr Arw.: 'does yno ddim ynd hen wr, Oct. 2,
1856; peth hawdd ydi dyud tos neb ynd ychunan yn gwbod sut
yr odd hi, ib.; ynd ran hyny (' but as to this' ), May 28, 1857, etc.
Onide, pron. ont, 'is it?'; nt in Carnarvonsh., Sweet, p. 411.

[31.] On behct, bet, see Ehŷs in Rev. Celt., vi, p. 57. Davies
(Dict., s.w.) twice mentions fecl Demet. = hyd Venedot.,
1 usque ad'. In Dosp. Ed., 543, 925, S.-W. med is given.

Sorne further details on divers nominal prepositions
(Zeuss, G. C.\ pp. 691-698), etc, are: S.-W. ym mysg =
N.-W. ym mhlith, Hughes, p. 33; N.-W. cyfyl = S.-W.
yn agos, Y Cyf Dyfyr (Euthin), p. 78; N.-W. ynghylch,
oddeutu = S.-W. obeutu, Y Traeth., iii, p. 14 (see YCymmr.,
viii, p. 159, and add obothtu, or bofftu, used at Neath).
Gwent. cera a dos o bothdy dy fisnis, minda dy fusnes =
Demet. gofala ani dy fisnis, Camb. Joum., iii, p. 248. This
word is too widely spread to be exphiined as an Anglo-
Welsh f orm introducing Eng. boii for deu in o-ddeu-tu, so we
nmst goback to o-bob-tu, o boptu, and ascribe ph, (Jfov p)
to an h developed by the accent. I have, however, no similar
examples except dathod, daffod: dattod?

L. Morris, Add. MS. 14,923, f. 133: S.-W. gwyddeneb =
N.-W, gyferbyn, pron. gwydderbyn (sic); godderbyn, see
Y Cymmr., vii, p. 235.

S.-W. serch in the sense of tros, er, Eichards, Dict. Liw
dydd, liw nos, ' by day, by night', Powel, Y Cymmr., vi,
p. 138.

[32.] Men, myn (' where') are frecpaently usecl in the poems
printed by Skene and in the Myv. Arch. Cf. B. of An.: men
na bei, men d nt eilassaf elein, Y God., Stanzas 43, 54, 1
etc. It is an oblique case of man, ' place', and has been
retained in the Breton dialect of Vannes (mnn, ' where').

April mii, 1888.

i 8k., ii, pp. 76, 79.


As to cynnagpwy, etc. (see YOymmr. f ix, p. 118-9), I have since found
in the Marchog Crwydrad (Y Brython, vol. v, 1863, p. 368 1 ): ein cyfoeth,
gynnag p'un a fo genytn ai ychydg ai llawer. In the same text occur:
ei dechreuad cynnaf, yn gynnaf, y dechreuad cynnaf p. 7 (cyntaf), of
which form two examples are found iu the Add. MS. 14,921, which
contains so many exaniples of cynnag pwy. Pwy hynnag and cyntaf
occur scores of times in the Marchog Cricydrad; but since the editor
states in the preface to his edition that he often introduced modern
Welsh orthography, one cannot tell whether gynnag, cynnaf were ordi-
narily altered by him into bynnag, cyntaf, or whether they really occur
only once or twice in the MS. The editor says (Preface, p . 2) " bod yr
ysgrifenydd wedi myned yn fynych i eithafion gwerinaidd y dafodiaith
hno [i.e., y Ddeheubartheg]"; the language, as far as the scanty remains
of dialect permit nie to judge, agrees more than that of any other text
I lcnow of with that of the Gwentian Add. MS. 14,921. Now theoccur-
rence of both cynnag pwy and cynnaf (cf. also ond cygynned ag y cywo
ef flas pechod, p. 369 2 ) in these two texts renders the suggestion I made
/. c. almost certain to me; viz., that pwy gyntaf and pwy bynnag were
mixed up, and that, in the Hmited district to which these two MSS. be-
long, cynna(f) and bynna, bynnag caused cynnag to be formed; cynnaf
itself was probably caused by cyn. Since, however, at present only
cynta pwy seems to be used, it remains a question whether the simi-
larity of cynna and bynna caused cynnag, cynta being afterwards intro-
duced instead of cynna(g), or whether cynta(f) and bynna(g) were
directy mixed up, and cynna(g) lived only for a certain time or in a
certain dialect, as long as or where cynna(f), for cyntaf, was used.

May 28th, 1888.


P. 264,1. 24. Llefydd is also the Southwalian form, and is in no way
parallel to torfŷdd, etc, as it is accented on the first syllable. It seems
to be formed after the analogy of such words as tre\ trefydd.

P. 269, 1. 27. Onithtar. In thevale of (lamorgan I have heard this
niade into cnfftar.

P. 278, 1. 19. Shoe. This word is one of the forms taken in Welsh
by the English "show", and we say in N. Eeredigion: 7 sio'e o bobol
' n edrach ar y sie, " there is a sliow (a sight = multitude) of people
looking at the show (the menagerie)". For sie o bobol we might also
Bay pwr o bobol, where we employ tlie English word " power" as it is
sometirnes used in Englisli.

1 Pt. iii, ch. 6; Reprint, p. 50, col. 1. - Jl>.; Rcprint, p. 51, col. 1.



By E. P.

A: see"MS. A".
A * * *: see "PA * * *"

Ab Iolo: = Taliesin ab Iolo. See " Cyfrinach, etc", " Iolo MSS."
abbrev.: abbreviatiou.

Add. MS.: = one of the collection of "Additional MSS." in the British
Museuin, which comprises (inter alia) the two great Welsh collec-
tions of the Welsh School aud the Cymmrodorion Society.

Add. MS. 14,931 (" Welsh School MS."): see " MS. E".
Add. MS. 15,055: see " W. Lleyu, Yocab."
Add. MS. 22,356 (" Cynnurodorion MS."): see " MS. <S ,M .
Add. MS. 31,055: see "Thos. Williams, Dr."
Ams.: see " IV Ams."
An.: Aneuriu: see "B. of An."
Angl.: Anglesea.
Apr.: April.
Araith Ieuan Brydydd Hir: 'the speech of Icuan Brydydd Hir (Hynaff,

who fi. 1440-1470. The copy cited is that iu Add. MS. 14,973.
Araith y Trwstan: ' the Awkward One's Speech'. The name of a long
poem by Siou Tudur (died 1602). The copy cited is that iu
Add. MS. 14,987.
Arch. Brit.: Edward Lhuyd's [cdias Lhwyd] "Archieologia Britannica,
etc, vol. i, Glossography" [all published] (xford, 1707, folio).
Note. The Welsh Preface (At y Cymry) of 6 pages, quoted more
than once, is unpaged.
Arch. Camb.: " Arch&ologia Cambrensis; the Journal of the Cambrian
Archaological Association", 1846, etc. (4 Series completed and a
5th in progress).
Ardudwy: This district comprises the littoral of Merionethshire

between the Mawddach and the Traeth Bach.
Arw.: see li Yr Arw."

At y C'ymry: ' To the Welsb.' See "Arch. Brit."
Aug.: August.

B: see " MS. B\ B.: see " R. B. Mab."
B. of An.: Book of Aneurn.

B. ofAn., God. (or YGod.): id., The Gododin. The original MS. (13th
cent.) was in the late Sir Thomas Philhpps' coliection at Middle


Hill, Worcestershire (now in the possession of his son-in-law,
Mr. Fenwick, of Cheltenham), and contains (1) the Gododin,
(2) the Gorchanau, viz., Gorchan Tmlficlch, G. Adebon, G. Cyn-
felyn, and G. Maelderw. The editions quoted are (1) that of Mr.
Skeue in his Four Ancient Boocs of Wales, vol. ii, pp. 62-107,
purportiug to represent the whole of the origiual MS. (Neither
lines nor stanzas are nurnbered in this editiou.) (2) Williams ab
Ithel's Y Gododn (Llaudovery, 1852, 8vo.), containing only the
Gododin without the Gorchanau , and edited from copies of the
aboye-named MS. and others. (Both lines aud stauzas are num-
bered in this edition.) Other editions of the whole Book are to be
found in the Myv. Arch. (lst and 2nd eds.), aud of the Gododin
aloue in Stephens' work, " The Gododin of Aneurn Gwawdrydd, 1 *
printed by the Cymmrodorion Society. See " Sk. (or Skene), ii."

B. of Carm.: This meaus the " Black Book of Carmarthen" (Hengwrt
MS. 11). The references are (1) to the pages of Mr. Skeue's
edition in qp. cit., vol. ii, pp. 3-Gl; (2) to the folios of the autotype
Facsimile of the MS., brought out by J. Gwenogvryn Evans
(Oxford, 1888), The MS. is of the late 12th aud. early 13th
centuries. See ''Sk. (pr Skene), ii."

B. of Herg.: This meaus (not the White, but) the Red Book of Hergest,
a MS. of the 14th cent. in the Library of Jesus Coll., Oxford.
The references are (1) to the MS numbered (not by folios, but)
by coumns; (2) to the pages of the Tcxt of the Mabinogion, ( tc.,from
the Hed Book of Hergest, edited by Professor Rhŷs and J. Gwen-
ogvryn Evans (Oxford, 1887): see " R. B. Mab."; (3) to the pages of
Skene, op. cit., vol. ii (pp. 218-308), where parts of the poetry in
this MS. (coll. 577-585 and 102G-105G) purport to be reproduced.
See " Sk. (or Skene), ii."

B. of S'l. Clad: The Book of St. Chad iu the Library of the Dean and
Chapter of Lichield Cathedral: a late 7th- or early 8th-ceutury
Irish MS. of the Latin Gospels, with Welsh marginal entries i>t
the 8th and 9th centuries. Those cited for the word maharun
(p. 2G7 siijj)-a) are of the (? later) ninth century, and by the
same scribe. (This book is often cited by Zeu.-s and otliers as
the Lichjuhl ('ni/i.r or Codi x Lichf.)

Note. The longer Welsh entries in bhis MS. purport to be re-
produced as an Appendix to Lih. Land. (<j. v.), pp. *271-4.

/;. of TaL: Book nf' Taliessin (Hengwrt .l/.s'. 17, a MS. of about the
middle of the 18th cent.). The references are to the pages of
Skene, op. cit., vol. ii, pp. 108-217, wliere this MS. purports to be
reproduced. See "Sk. (or Skene), ii", and "MS. V, W".

■- Barddas: or,;i Collection of Original Docuincnts, illustrative of the
Theology, Wisdom, and Usagesof the Bardo-Druidic Systein of the


Isle of Britain; . . . . by the Kevd. J. Williams ab Ithel. For

the Welsh MSS. Society, Llaudovery, etc, 1862" (vol. i).
Note. Only part of vol. ii published.
Bas-Cornouaillais: = the Breton dialect of Lower (i.e., Western) Cor-

nouaille, comprising the S.W. of the Dpt. of Finistre.
Bas Vanu.: Bas-vannetais, i.e., the Breton dialect of the lower (=:

western) part of the Pays de Vannes, including (roughly speahing)

the couutry betweeu the rivers Scorff (on the E.) and Ell (on the

W.), in the Dpt. of Morbihan.
Batz: means the Bourg de Batz, near the mouth of the Loire (De'pt. of

Loire-Dfrieure), not the Ile de Batz, in the Pays de Le'on (Dpt.

of Finistre). Ernault has written on the isolated Breton dialect

of Batz in Rev. Celt., iii.
Bed,: see " Y Bed. v
Beitr.: " Beitrge zur cymrischen grammatik. (einleitung und voca-

lismus.) der philosophischen facultt der universitt in Leipzig als

dissertation zur erlangung der philosophischen doctorwrde eiu-

gericht von Max Nettlau, aus Neuwaldegg iu Niedersterreich.

Leipzig, mrz-april 1887."
Bibl. Bodl.: (see " Cambrica") Bibliotheca Bodleiana.
Bodl.: = one of the Bodley MSS. in that Library.
Bown o H.: Bown o Hamtwn; i.e., the Welsh version of Sir Bevis of

Hampton, preserved in the Llyfr Gwyn Rhydderch (see under " Ll.

Gw. PJi.") and the Red Booh of Hergest.
Brec.: Breconshire, alias Brecknockshire.
Bret.: see " Rev. de Bret."
Brit.: see "Arch. BriO

C: (1) Cymru. See " Can. y C", " Llyfr. >j C", " S. C."
C: (2) Cymreig. See " P. C"
C: see " MS. C". C 2: see " G. 6'. 2 ", " Zeuss".
Cab. few.T.: Caban fewythr Tomos. The work cited is: " Aelwyd

fewythr Robert, neuhancs Caban feuythr Tomos, gan William liees

[Gwilym Hiraethog], Dinbych [Denbigh], 1853."

Note. Not to be confounded with Caban fewyrth Twm, the

Welsh translation of Uncle Toni's Cabiu (Cassell, London, 1853,

and other editions).
Caledfryn, Gramm. 2: = the second edition of the Grammadeg Cymreig

of Gwilym Caledfryn CWilliam Williams), 1870.
Calig.: see " MS. Calig."
Camb.: Cambrensis: see " Arch. C?/ti."
Cambr. (1): Cambrian: see " Cambr. Journ. v

Cambr. (2): Cambrica [by Whitley Stokes], in the " Transactions of
the Philological Society for 1860-1", comprising, " i. (pp.
20-1-232), The Welsh Glosses and Yerses in the Cam-


bridge Codex of Juvencus" [9th cent.]; " u. The Old
Welsh Glosses at xford, Bibl. Bodl. Auct. F. 4-32": viz.,
" 1 (pp. 232-4), Glosses on Eutychius" [since proved to be
ld-Breton]; '■ 2 (pp. 234-6), Glosses on Ovid's Art of
Love"; "3 (pp. 236-7), British Alphabet"; "4 (pp. 237-8),
Note on Measures and Weights"; [5] (pp. 238-249), [The
glossed portion of] "Bodl. 572" [since proved to be Old-
Cornish]; "m (p. 249), [Some of] The Middle-Welsh
Glosses in Cott. Vesp. A. xiv . . . . fo. II 1 " [the last 3 from
fo. 13 b ]; and iv, " Addenda et Corrigenda'' (pp. 288-293).
Cambr. Joum.: " The Cambrian Journal, published under the auspices
of the Cambrian Institute": 12 vols. (last one unfiuished), Tenby
(printed), 1854-1865.
Campeu Charlymaen: ' The Exploits of Charlemagne.' Of the Welsh
version of this work two MSS. have been published. For the one
quoted underthe above title, see "LL Gw. /." For the other, see
" Yst. de Cnr. Mag." Both MSS. are of the 14th century, but
represent different editions.
Can. y C.: " anwyll y Cymru [by Rice Prichard, Vicar q Llando-
very; 1579-1644] yubedair rhan, Llundaiu, 1672" (2nd edu., 8vo.).
Capella Glosses: = " M. Cap. 1 ', q. v.
Card.: Cardiganshire.
Carm.; see " B. of Carm."
Carn.: Carnarvonshire.
Celt.: see " Rev. Celt."
cent.: century.
Chad: see " B. of St. Chad."
Chr.: see " Y Drych Chr."

Chwedlau V Doethion: ' The Wise Men's Sayings'; a collection of
Welsh proverbial Tripleta (oldest MS., about 1330-1350, in Jes.
Coll, O.foii., MS. No. 20). Tie editiou of these referred to is
tlmt in the loh MSS. (pp. 251-9), q. v.
cf.: conj't r.
Ch.: chapter.

Cl., Cleop.: = " MS. Cleop.", 7. v.
col.: colunm; coll.: coluinns.
comm.: of common gender.
Corn.: Cornish. See also " O.-Corn." '

Corn. Vocb.: The ancient (Latin-) Cornish Vocbulary, in the begiu-
ning of C'o/t. Vesp. A. xiv (very early 13th cent.; transcribed from
a considerably older MS.), printed in Zeus8' Grammatica Celtica,
lst ed., pp. LlOO-1124, 2nded.,pp. 1065-1081; and, in alphabetical
order, by Edwin Norris in vol. ii of liis Ancient Cornish Drama
(Oxford, at the University Press, 1859, 8vo.), pp. 819-432.


Cott.: = Oue of the MSS. of tlie Cottonian collection in tlie British

Museum. See " MS. Calig.", etc. " MS. A, etc."
Cyf.: see " Y Cyf., etc"
yfrinach Beirdd Yn. Pr. (Abertawy, 1829): Cyfrinaeh Beirdd Ynys

Prydain (edited by the late Iolo Morganwg, and brought out by his

son, Taliesin al> Iolo), Abertawy [Swansea], 1829.
Cychgr.: see " Y Cylchgr."
Cymmr.: Cymmrodor; see " 1' Cymmr."
Cymmrodoron Society (Works published by): see " B. of An.", " Hgt.

MS. 202", "L. Glyn Cothi", " MS. Tit. D xxii", Yst. de Car.

Magno And cf. " MS. S".
Cymr.: (1) Cymric (i.e., ' Welsh'), see " Stud., etc."
Cymr.: (2) Cymreig; see " Y Gicron Cymr."
D: see"MS. J D".
d.: died.
D. S. Evaus, W. Dict.: " A Dictionary of the Welsh Language, by

the Revd. D[auiel] Silvan Evans" (Parts i, contaiuing " A", and

ii, containing " B", are out), Wm. Spurrell, Carrnarthen, 1887-8.
Id.: see "Gwallt. Mech.", "Llyth.", "Llyfr. y C.", "Marchog Crwydrad?
Dares Phrygius. The Welsh version quoted is the one in Cleop. B. V.
Daf. ab Gvv.: Dafydd ab Gwilym. The edition of his poems cpuoted

here is the frst one, by Owen Jones (Myfyr) and William Owen

(afterwards Dr. W. O. Pughe), London, 1789, 8vo.
Davies: = The Revd. Dr. Joliu Davies of Mallwyd (1570-1644). See

" Ll. y Eesol."
Davies, Dict.: = IIis Welsh-Latin and Latin-Welsh Dictionary, entitled

" Antiqua3 Linguse Britannicse . . . . et Lingua Latiuse Diction-

arium duplex, etc, Londini, etc, 1632" (4to.).
Davies, Gramm.: The^Vs<edition of his Gramtnar, intituled: " Antiquse

Liugua Britannicse, etc, Rudimenta, etc." (London, 1621, 12mo.).
Note. The second (and last) edition (Oxon, 1809) is also ouce

quoted co nomitie (p. 271, supra).
De r Urgence, etc.: " De l'urgence d'une exploration philologique en Bre-

tagne, ou la langue bretonue devant la science. Extrait des

Mmoires de la Socit cV Emidation des Ctes-du-Nord ,, (St. Brieuc,

1877, 8vo.; pp. 18) [par Emile Ernault].
Dec.: December.
Demet.: Dr. John Davies' occasional abbreviation for Demetie, Demetas,

Demetic, etc, ' the Demetians, Demetiau', i.e., the people or (in)

the language of Demetia (Dyfed). The otherform is "Dimet.", q. v.
Denbigh: = Denbighshire.
Dpt.: Dpartement.
Dict.: Dictionary. See " Davies", " D. S. Evans", " l'A * * *", " Owen

Pughe", " Richards", " Sp."


Dimet.: Dirnetian, etc. Davies uses " Dimetian" to indicate ' S.-Welsh';
it is less improperly used by most writers to nclude the dialects o
Pembrokeshire, Carmartheushire, and Cardiganshire, embraciug
the old divisions of Dyfed, Ceredigion, and all Ystrad Tywi but
Gower, which is now in Glamorganshire.

Dosp. Ed.: "Dosparth Edeyrn Davod Aur; or the Ancient Welsh
Grammai; . . . . by Edeyrn the Golden-tongued, etc, etc, with
English translatious and notes, by the Rev. John Williams Ab Ithel,
etc. Pubd. for The Welsh MSS. Society, Llandovery, 1856."

Note. The " translation" of the Grammar, quoted by Dr.
Nettlau, is virtually a uew Welsh Grammar by the Translator (see
his Pref., p. xv).

Dr. Thos. Williams: see " Thos. Williams, Dr."

Drych, etc.; see " Y Drych, etc."

Dyfed: The Welsh for Dimetia. See " Demet.'\ " Dimet."

E: see " MS. E."

E. Glam.: Eastern Glamorganshire, comprising all the couuty east of a
line drawn from Merthyr Mawr (near Bridgeud) to Aberdare
(Camb. Journ., iii, p. 2JT).

E. Evans: (The late) Evander Evans. See " Stud., etc."

E. Lhuyd: Edward Lhuyd. See " Arch. Brit."

e.g.: exempli grati.

Early Engl. Pron.: Early Engsh Pronuncation, by Alexauder J. Ellis.

Ed.: see " Dosp. EdP

VA.: edited; edn.: edition; eds.: editious.

, , , ' \ see " Early Enal. Pron."
Engl.:) * J

Emj.-W. Dict.: see "Sp., etc."

Ergyng (p. 270, n.): = Erging, now the Deanery of A rchenfield, compris-

ing the S.E. portion of Herefordshire E. of Wye.

Ernault: = M. Emile Ernault. See " Batz", " De l'Urgence, etc."

Evans: (1) see " D. S. Evans, etc"

Evans: (2) Evander Evans. See " Sttid., etc."

Evans: (3) J. Gweuogfryn Evans. See ' B. of CannJ', " R. B Mab."

Exli-i\: Exercises. See " ltowlaud, Exerc."

F: see "MS. F.

f.: folio (of a MS. or book).

Feb.: February.

fem.: femininc

f'eic.: see " Cab.fcw. T."

if.: fulios (of a MS. or book).

.//'.: see " Hanes //.//'."

//.: joruit.

Flint: Flintsbire.


freq.: freqventer. (This abbrev. is always quoted from Zeuss, G. C. 2 ).

G.: see "Iolo G." G: see "MS. (?".

G. C. 2 (also " Z 2 ", " Zeuss'').= The second edition of Zeuss' Grammatica

Celtica, by Ebel (Berlin, 1871, 4to.).
GL, gl.: Glosses, gloss. See " Juv. GL", " Camlrca", " M. Cap."
Glamorgansh.: \ Glamorgan or Glamorganshire. See " E. Glam.", " W.
Glam.: Glam."

God. (also Y God.): The Gododin. See " B. of An."
Gramm.: Grammar. See " Caledf ryn", " Davies", "Rowland","Spurrell."
Greal: see " Y S. Greal."

Gw.: (1) see " Ll. Gic. Rh. v Gw.: (2) see " Lliccr, etc v
Gwallt. Mech., Worl-s; " Gwaith y Parch. Walter Davies, A. C.

(Gwallter Mechain), etc." (Ed. by the Revd. D. Silvan Evans,

3 vols., Carmarthen, 1868, 8vo.)
Gweith.: see " Y Gweith."

Gwenogfryn: ) (= J. Gwenogfryn Evans), see " R. B. Mb", " B. of
Gwenogvryn: ) Carm."
Gwent.: Gwentian, i.e., of Gwent or in its dialect. Used conventionally

of the dialect of Gwent and Morganwg, i.e. (roughly speaking)

Monmouth- and Glamorgan-shires. (Gwynllywg, between the Usk

and Rumney rivers, though cow in Monm., was anciently in Mor-

ganwg. The name is now corrupted into Giuentllwg.)
Gwyl.: see Y Gwyl.

H: see " MS. E". H.: see " Bown o H
Hanes y ff., 1677: " Y Ffydd Ddi-fvant. sef, Hanes y Ffydd Gristi-

anogol, etc." (3rd ed., Oxford, 1677, 8vo.), by Charles Edwards.
Hanesion o'r Hen Oesoedd, 1762 [alias 1872): (quoted on pp. 262, 264,

and 281 supra).
Harl. MS.: = Harleian MS. 1796, a MS. of the Latin codification of

the Welsh Laws. See " Latin Laws".
Harl. MS. (with number following): = One of the Harleian collection

of MSS. in the British Museutn.
Harl. MS. 958: see " MS. T, and cf. " Ll. Gw. Rh.", " Hgt. MS. 202."
HarLMS. 4353: see " MS. F".
Hengwrt MSS.: see " Hgt. MS.", etc, " Wms., Hgt. MSS., ii." (For

various other Hengwrt MSS. see under " B. of, etc", " Latin

Laws", "Ll. Gw. Rh.", "MS. A, etc", " W. Lleyn".)
Her. Vis.: " Heraldic Yisitations of Wales and the Marches in the time

of Queen Elizabeth and James I, by Lewis Dwnn" (edited for the
Welsh MSS. Society by the late Sir S. R. Meyrick; Llandovery,

1846, 2 vols., imp. 4to.). See " Llyfr Achan".
Herg.: see " B. of Herg."

Hgt. MS. (Latin Laws): = Hcngwrt MS. 7. See " Latin Laws".
Hgt. MS. 57: The unique fragment quoted hence (p. 279 supra) occurs


q the niiddle of the MS., on three pp. whence the original writing
has been obliterated.

fgt. MS. 202: = A Fragment from Hengwrt MS. 202, edited in Y
Cymmrodor, vol. vii, pp. 89-154. The folios by which this MS. is
cited supra are (1) not those of the MS. rolumc, but of the frag-
ment stitched into it, which has nothing to do with the rest of its
contents, (2) quoted by the numbers as first printed, subsequently
found tobe wrongly read, and corrected in Y Qammr. } vii, pp. 20-1-6.
This fragment is in the same style of hand as the Mabinogion in Ll.
Gw. Rh. (Hgt. MS. 4), q. v. and Harl. MS. 958 ("MS. T"), q. v.
All are of late 13th or early 14th century date.

Hgt. MSS.: see " Wms., Hgt. MSS., ii."

Hibb. Lect., Rhŷs: " The Hibbert Lectures, 1886. Lectures on the Origiu
and Growth of Religion, as illnstrated by Celtic Heathendom, by
[Professor] John Rhŷs .... Londou, etc, 1888."

Hom.: see " Three, etc."

Homil.: " Pregethau a osodwydallan trwy awdurdod i'w darllein ymhob
Eglwys blwyf a phob capel er adailadaeth i'r bobl annyscedig.
Gwedi eu troi i'r iaith Gymeraeg drwy waith Edward James"
(London, 1606, small 4to.). [The first edition of the Welsh transla-
tion of the Homilics.~\

Hope, Poems (1765): The work from which these poems are quoted
is: " Cyfaill i'r Cymro; neu, Lyfr o Ddiddanwch Cymhwysol, Ei
dowys Dyn ar Ffordd o Hyfrydwch: Yn Ganeuon, a Charolau, ag
Englynion hawdd yw deuall. O waith Prydyddion Sir y Flint, a
Sir Ddimbech. O Gasgliad W. Hope, o Dre Fostyn. Caerlleon
[Chester]: Argraphwyd gan W. Read a T. Huxley yn y Flwyddyn
1765; ag ar werth gan W. Hope, yn Sir y Flint."

Note. William Hope of Mostyn (in Welsh Tre Fo.<;ty),Flintshire,
was one of the eight authors of the poems contained in this book.

Hughes (1823): == "An Essay, on the ancient and present state, of the
Welsh Language: with particular reference to its dialects. Being
the subject proposed by the Cambrian Society, for the year 1822.
By John Hughes, author of Horat Britanncs0' (8vo., London, 1823).

T: see " MS. 7". I.: see " Sal, N. T, Gwel. /."
Introd.: Tntroductio.

ib.: ibidem; id.: idem.

Iolo G.: Iolo Goch (died after 1402). Tle is sometimes cited as " Iolo",
sometimes as " I. G.", in the works of Dr. John Davies of Mallwyd.

lolo MSS.: " Iolo M(i>t)isrr//,ts. A Selection of Ancient Welsh Mumi-
scripts, in Prose and Yerse, from the collection made by tlie late
Edward Williams, Jolo Morganwg . . . . l>y liis son, the late
Taliesin Williams (Ab Toh) of Merthyr Tydfil Published for Tlc
Welsh MSS. Society, Llandovery .... 1848.


J: see " MS. J".

Jes. ColL, 0x0%., MS.: = One of the MSS. in the Library of Jesus
Coliege, xford.

J. Gwenogvryn Evans: see " B. of Carm", " R. B. Mab."

Jan.: January.

Jur. Gl.: Juvencus Glosses. See " Cmbrica".

Kuhn's Zeitschr.: Kuhn's Zeitschrift.

l'A * * *, Dict.: " Dictionnaire franois-breton ou franois-celtique du
dialecte de Vannes. Enrichi de thmes .... par Monsieur
L'A * * *.
(Supplment considrable aux Dictionnaires franois-
bretons) - ', Leide, 1744, 8vo.

L: see " MS. L".

L. Glyn (or Gl.) Cothi: " Gwaith Leicis Giyn Cothi .... Oxford, for
the Cyramrodorion .... 1837'' (2 vols., 8vo., but with onepagina-
tion runniug throughout). L. G. C. died after 1486.

L. Morris: Lewis Morris (Llywelyn Ddu o Fn; d. 1765), brother to
Richard and William Morris.

Latin Laics: The Latin codifications of Welsh Laws, printed in Aneurin
Owen's Ancient Laws and Institutes of Wales, frora the MSS., and
at the pages of vol. ii of the 8vo. editiou, raentioned below:
Hengwrt MS. 7 (early I3th cent.); pp. 749-814.
MS. C'ott. Vesp. E. xi (early 14th cent.); pp. 814-892.
Harleian MS. 1796 (13th cent.); pp. 893-907.
N.B. Dr. Nettlau's paginal references are to the folio edition.

Lect.: see " Hibb. Lect."

Lect. 2 , Rhŷs: " Lectures on Welsh Philology, by [Professor] John Rhŷs.
Second edition .... London, 1879."

Lewis: see " L. Morris", " L. Glyn Cothi."

Lhuyd: see " E. Lhuyd", " Arch. Brit."

Lib. Land.: " The Liber Landavensis, Llyfr Teilo, or the Ancient
Register of the Cathedral Church of Llandaff .... Published for
The Welsh MSS. Society, Llandovery: . . . . 1840.*' (Original MS.
written in about 1132. R. W. Haddan.)

Ling.: see " Mm. Soc. Ling."

LL.: The Leabhar Laigneach, or Book of Leinster, a MS. of the 12th
century, in the Library of the Royal Irish Academy.

Ll. Gw. Rh.: Llyfr Gicyn Rhydderch, ' Rhydderch's White Book',
(Hengwrt MSS., Nos. 4 and 5). This consists of parts of two
separate MS. collections bound together, the older one (in the same
style as the Fragment from Hengwrt MS. 202 and Hetrl. MS. 958)
of the late 13th or early 14th century, and containing the " Mab-
inogior" (all but Rhonabwy), the other of somewhat later date.
The " pages" to which reference is made as those of this MS. are
those of the Revd. Canon Williams' Selections from the Henrpnt


MSS., vol. ii, pp. 1-284 (Parts IV and V of the whole work, Part
VI being unpublished), in which portions of the MS. (all, but Boicn
o Hamtum, taken from the niore modern part) pnrport to be repro-
duced. The followiug are the works printed by Canon Williams
from his transcripts of this MS.:

Campeu Charlymaen, Hengwrt MSS., vol. ii., pp. 1-118.

Bown o Bamtwn ,, 119-188.

Purdan Padrc ,, 189-211.

Buched Meir Wyry 212-237.

YSeithPechawtMarwawl,, 237-242.

Euangel Nicodemus ,, 243-250.

YGroglith 250-266.

Hanes Pontius Pilatus) 9fi

Historia Judas } " " " " /b '"^ 4 -

Arwydon cyn dydbrawt (o weih Lywelyn vard) ,, 274-275.

Prophwydolyaeth Sibli Doeth 276-284.

The rest of the contents of Canon Williams' 2nd volume are

referred to supra as from " Wms., Hgt. MSS., ii", q. v.
Ll. y Resol.: " Llyfr y Resolusion .... Wedi ei gyfieithu yn Gymraeg

gan y Dr. I[ohn] D[avies] er lls i'w blwyfolion . . . ." (2nd ed.,

London, 1684, 8vo.)

A translation of the English work by Father Parsons, the Jesuit.
Lleyn: see ' AV. Lleyn".
Llicer Gw. Gy)'., 1586: Ll'wer Gweddi Gyffredin, etc London,

1586, 4to.
Llyfr Achau: (' Pedigree-Booh'). This i-efers to the one by Hopkin ab

Eignon of Brecon (1602), printed in " Her. Fẃ." (q. v.), vol. ii.
Jjifr. y C.: Llyfryddiaeth y Cymry, or Cambrian Bibliography, by the

Revd. Wm. llowlands (Gwilym Lleyn). Edited and enlarged by the

Revd. D. S. Evans (Llanidloes, 1869, 8vo.).
Llŷn: = Ijleyn.
Llyth., D. S. Evans: " Llythyraeth yr Iaith Gymraeg, gan [y Parch.]

D. Silvan Evans. Caerfyrddin [= Carmarthen], W. Spurrell, 1861."
Loanwords: " Welflh Words borrowed from Latin, Greek, and Hebrew"

(by Professor Rhŷs), in Arch. Camh., 4th Series; vol. iv (878)i

pp. 258-270 and 355-365; vol. v (1874), pp. 52-9, 224-232, and

71/: see "MS. M".
M.Cap.: ="The Old Welsh (losses on Martianus Capella['s h

Nuptiis Philologia ci M< ^^n■ii~\ ,, , edited (and numbered) by Whiey

Stokes from the original in MS. C.C.C.C. 153 (8th cent.), in Arch.

Camb. for 1873, No. 13 of 4th Series [vol. iv], pp. 1-21.


Mab.: Mbinogon. "The Mbinogion froin the Llyfr Coch o Hergesf,

etc., by Lady Charlotte Guest" (3 vols., Llandovery and London,
1849). See also " B. of Herg.", " Ll. Gw. Eh.", " R. B. Mab."

Note. (Where " Mab." follows a quotatiou from Zeuss, it really
forrns part of the quotation from that work.)

Marchog Crwydrad (Reprint): " Y Marchocj Crwydrad: Ilen Ffnglith
Gymreig. Cyhoeddedig dan olygiad y Parch. D. Silvan Evans.
Tremadog: R. I. Jones [Alltud Eijon]; Caerfyrddin: W. Spurrell,
1864." (8vo.)

Note. This was first published in parts in Y Bryihon, vol. v
(1862-3), pp. 1-17, 138-153, 257-267, and 361-374, whence Dr.
Nettlau quotes, but concordances are appended to the above
edition wherever the pagiuations differ. Chapters 1 to 6 of Book i
are similarly paginated (pp. 1-17) in both editions.

masc.: masculine.

Me'm. Soc. Ling.: Mmoires de la Socit de Linguistique de Paris.

Menevia: used in an article quoted from Cambr. Journ., iii, iv, for
Mynwy, ' Monmouthshire'. Properly it means Mynyw, 'St. David's'.

Merioneth: Merionetbshire.

Mid.-I.: Middle-Irish. See " Three, c/c."

Mid.-W., Middle W.: Middle-Welsh.

Monm., Monmsh.: Monmouthshire.

Morganwg: see " Gwent."

Morris: see u L Morris", " R. Morris".

MS.: Manuscript. See " Add. MS.", " Harl. MS", etc, " IJr/t. MSJ',
etc, " Latiu Laws"; and the following entries.

MS. C.C.C.C: (see "M. Cap.") one of the MSS. in the Library of
Corpus Christi College, Cambridge.

ne of the MSS. designated Caligula, Cleopatra,

Titus, or Vespasian, respectively, in the Cottoniau

"i collection at the British Museum. (For various of

these MSS. see under " Dares Phrygiis", " Latin

Laws n , "MS. .1, etc")

MS. Tit. D xxii: A MS. of the early 15th cent. (part is dated 1439),
some of which (fos. 1-19, comprising A Description of the Day of
Judgement) was edited, with notes, by Prof . Thos. Powel in Y Cymm-
rodor, vol. iv, pp. 106-138.

MS. A, etc: These MSS. are those mainly used by Aneurin Owen for
the Welsh part of his Aucient Laws and Institutes of Wales (London,
1841), and distinguished by him for critical purposes by the suc-
cessive letters of the alphabet. The pages appended to the citatious
of the MSS. supra are those of thefolio editionof the printed work,
which was also published simultaneously in two large vols. 8vo.
The following is a list of the MSS. in question:

MS. Calig.:
MS. Cleop.(orCl.)
MS. Tit.:
MS. Vesp.:


[I. MSS. of Venedotian Code.] Dates of MSS.

MS. A: (Part of) Hengwrt MS. 26 (" Llyfr du o'r PTaera") (about 1241)
., J5: Cott. Titus D. II. - - (13th cent.)

,, C: Calnj.A.lU. - - (about middle of I3th cent.)

(Parts of two distinct MSS. bound together.)
,, D: Hengwri MS. 311, and (3 pp. misbound) part of

Henywrt MS. 8 (" Llyfr Tg") - (Do. 14th cent.)

E: Add. MS. 14,931 (" Welsh School MS.") (middle of 13th cent.)

F: Hengwrt MS (perhaps No. 5 of Hgt. MS.

39; see next entry).
G: Hengwrt MS. 39 (No. 2 or No. 5?).

Note. This MS. vol. comprises five separate MSS.:

(1) Dimetian Laws, MS.,pp. 1-25.

(2) Yenedotian Laws, pp. 26-50.

(3) Llyfr Cynghawsedd (also in MS. B), pp.

52-71 - - - - (early 13th cent.)

(4) Llyfr Cynog, pp. 73-76.

(5) Venedotian Laws, pp. 76-119.

Nos. 1-4 are intituled "Cyn'\ and No. 5 " Adcyu"..
H: (Part of) Hengwrt MS. 26 (see above) - - (16th cent.)

[II. MSS. of Dimetian Code.~]
MS. /: Hengwrt MS. 19 (" Beta 19") - (about middle of 14th cent.)
J: Jesus Coll. Oxon. MS. No. . . . (late 14th or early lth cent.)
K: (Part of) Hengwrt MS. 18 ("Ealan.") - (about 1469)

(Said to contain poems in the autograph of L. Glyn Cothi.)

L: Cott. TitusD. IX - (late 13th or early 14th cent.)

., . , t* , j, (before end of I4tli cent.

M-.HeugirrtMSAl ("Beta47") - | v W.WE.W.)

N: (Partof) Hengwrt MS. 312 ("Beta") - (Uth cent.)

: Do. (" Bedu") - (about 14th cent.)

/\- (Part of) Hengwrt MS. 6 (" Befol") - (early 15th cent.)

Q: A Wynnstay MS. (burnt in 1858) - (about 1401)

r. tt iro oo /nro r i 7,ri ,, (mddle of 14th Ceilt.

R: Hengwrt MS. 23 (MS. Maredudd Llwyd) y T T1/ TTr

,, S: Add. MS. 22,356 (" Cymmrodorion MS.") (about 15th cent.)

(Written in South Cardiganshire.)
T: Harleian MS. 958 - - (late 13th or early 14th cent.)

(In the same style of writing as the Mbinogion of Hengwrt
MSS. 4 and 5, and the Fragment in Id. 202.)

[III. MSS. of Gwentian C'ode.]
MS. U: Hengwrt MS. 31 (" Morg.") - (about middle of Mtli cent.)
V: Harleian MS. 4353 - - - (13th cent.)

(In siine style, if not band, as B. of 'l'al. and fengwrt MS. . f >9.)

I see " Sal., N. 7Y


MS. W: Cott. Cleop. A. XIV. - - - (13th cent.)

(n same style as MS. V.)
X: Cott. Cleop. B. V. - (about middle of 14th cent.)

Y: MS. of (the late) Neath Literary and Philosophical Society

(lost since before 1860. 2?. P.) (middle of 14th cent.)

Z. (Part of) Hengwrt MS. 6 ("Pomf.") - (about 1480)

Note.- -Some of the above MSS. contain not only the three
Codes, but various parts of the Anomalous Laws, printed in vol. 2
of the 8vo. edition. The three other MSS. used by Owen for the
Welsh Laws are notcited by Dr. Nettlau; for the MSS. of the Latin
Codes used by Owen, see " Latin Laics".

Myv. Arch.: "The Myvyrian Archaiology of Wales", 3 vols., 8vo.,
London, 1801. A new edition in one vol., published by Gee. Den-
bigh, 1870.

-V: see "MS. N\

n.: note.

N. T, Sal.:

N Test., Sal.

N. W.: North Wales; N.-W.: North-Welsh.

Nov.: November.

O: see "MS. 0".

O.-Corn.: Old Cornish. The words so designated are from the Bod-
lcy MS. 572. See under " Cambrica".

Oct.: October.

op. cit.: opere citato.

Ovid Glosses: see under " Cambrica".

Owen: =: Aneurin Owen's Folo edn. of the Welsh Laws. See " Latin
Lau-s", " MS. A, etc."

Owen Pughe, Dict.: " A Dictionary of the Welsh Language, etc.
Second edition, by W. Owen Pughe, D.C.L., etc.; Denbigh,
Thomas Gee, 1832" (2 vols. 8vo). The quotations are made from
this edition: the lst (1803) bears the name of " William Owen"
(Dr. P.'s then name); the so-called 3rd edition (Gee, 1866) is
virtually a different work, based on Pughe, by Mr. R. J. Pryse.

P: see " MS. P".

p.: page; pp.: pages.

P. C: Punch Cymreig (Holyhead, 1858).

Phil.: Phihlogy. See " Lect. r \ " Stud., etcP

pl., plur.: plural.

Powel (or Powell): = Professor Thos. Powel, see " B. of An."\ " MS.
Tit. D 22", " Yst. de Car. Mag. n

Powis. or Powys.: Powysian; i.e., belonging to the dialect of Powys.
(Used to include all Montgomeryshire and the S.E. parts of Flint-
shire and Denbighshire.)

x 2


Pr.: see " Cyfrinach, etc."
Pref.: Preface.

Pughe: = " Owen Pughe'', q. v.
pubd.: piblished.
Q: see " MS. Q".
q. v.: quod riilc.
R .- see " MS. R'\

R.B.Mah.: Red Booh Mabinogion, i.e., " Y Llyvyr Coch o Hergest, Y

Gywol I, Y Mabinogion, etc. The Text of the Mabinogion and

other Welsh Tales from the Red Book of Hergest, edited by John

Rhŷs .... and J. Gwenogvryn Evans, Oxford, 1887."

R. Morris: Richard Morris (d. 1779), brother to Lewis Morris and

William Morris.
Resol.: see tl Ll. y ResoU'

Rev. Celt.: Rerue Celtigue, Paris, 1870, etc.
(8 vols. and part of a9th out).
Rev. de Bret.: Revue de Bretagne et de Yendc.
Rh.: see " Ll. Gw. 2A."
Rhŷs: Professor John Rhŷs, See "Hibb. Lec', " Lect.-\ "Loanwords",

Richards, Dict.: " Antiquse Linguse Britannicse Thesaurus . . . . by
Thomas Richards, curate of Coychurch" [Wallic Llangrallo, Gla-
morganshire], Bristol, 1753.

Note. This is the first edition of the work, and that from which

the quotations are made.

Rowland, Gramm*: "A Grammar of the Welsh Language . . . . by

[the lateRevd.] Thomas Rowland,4th edn. . . . Wreiham [1876]."

Rowland, Exerc.: "Welsh Exercises' ! (by the same), Part I. Bala,

1870 (all published).
S: see " MS. ST\ S.: Seint, Sant [' Iloly 1 ]: see " Y S. Gr."
S. C. or S. Cymru: " Seren Cymru. Newyddiadur Teuluaidd Pythef-

nosol" (Caerfyrddin [Carmarthen], 1856-1860).
S. Gomer: Scren Gomer, Caerfyrddin [Carmarthen], 1814, etc.
s. v.: sub voce; s. w.: sub vocibus.
S. E.: South-East; S. W.: (1) South-West.
S. W.: (2) S. Wales: South Wales. S.-W.: South-Welsh.
Sal., N. T.: Salesbury, New Testament; i.e. f "Testament newyddein
Arglwydd Eesu Christ. Gwedy ei dynnu, yd y gadei yr ancyfiaith
'air yn ei gylydd or Groec a'r Llatin, gan newidio ffurf llythyreu
gairise-dodi. Eb law hynny y mae pob Gair a dybiwyi y vot yn
andeallus, ai o ran Llediaith y 'wlat, ai o ancynefinder y deunydd
wedi ei noti ai eglurhau ar 'ledemyl y tu dalen gydrychiol."
(By William Salesbury, London, 15G7.)
Notc A reprint published a1 Carnarvon in 1850 (8vo.).


Sal., N. T, Gioel. I.: = id., Giceled'ujaeth Ieuun, .c., 'The Bookof Reve-
lation' in Salesbury'sabove-named work (translated by John Huet,
a resident and incumbent in the neighbourhood of Builth).

Sanskr.: Sanskrit.

Sept.: September.

seq.: sequentia.

ser.: series.

Sk. (or Skene), ii: Skene, vol. ii; i.e., the second volume of " The Four
Ancient Books of Wales, .... containing the Cymric Poems
atti'ibuted to the Bards of the Sixth Century, by William F. Skene,
Edinburgh .... 1868" (contaming the Welsh Texts of the Blach
Booh of Carmarthen, and Boolcs of Aneurin and Taliessin, and >ome
of the poetry from the Rsd Booh of Herg< st). See " B. of An.",
" B. of Carm.", " B. ofHerg.", " B. of Tal?

Soc.: see " Mm. Soc. L'ing."

Sp., Dict. 3: " A Dictionary of the Welsh Language .... by William
SpuiTell. Srd Editiou, Carmarthen, 186G."

Sp., Eng.-W. Dict. 3: " An English-Welsh pronouncing Dictionary
.... (by) William Spurrell." (Third Editiou, Carmarthen, 1872.)

Sp., Gramm. 3: " A Grammar of the Welsh Language, by William Spur-
rell. Third Edition, Carmarthen .... 1870."

St. Chad: see " of St. Chadr

Stokes: = Whitley Stokes. See " Cambrica", " M. Cap/\ " Three Mid.-
Irish Homilies", " Toga Troi (LL.)'\

Stowe MS.: ne of the collection of the MSS. formerly at Stowe, since
the Earl of Ashburnham's, and now (all but the Irish and a few
other MSS.) in the British Museum.

Note. The numbers cited are those, not of Charles O'Conor's
Bibliotheca MS. Stowensis (2 vols., 4to.),but of the " Sale Catalogue"
used at the Museum pendingthe completion of an official Catalogue.

Stud. in Cymr. Phil.: Evander Evaus' Studies in Cymric Philolojy,
priuted, with a continuous numbering, in ihtArch. Camb., 4th Ser.;
No. i in vol. iii (1872), pp. 297-314; No. ii in vol. iv (1873), pp.
139-153; No. iii in vol. v (1874), pp. 113-123.

sup.: supra.

superl.: superlative.

Sweet: = Dr. Henry Sweet's Siiolcen North-Welsh, printed iu the Trans-
actions ofthe (London) Philological Society for 1882-4.

T.: (1) see "Sal., N. T"; T: (2) see " Cab.few. T."
T: see " MS. T".
Tal.: see " B. of TaW

Thos. Williams, Dr.: Dr. (sometimes called " Sir") Thomas ^illiams (or
Ap Wiliem"), of Ardde 'r Myneich (now YrArdda), Trefriw


(Ji. 1573-1620), quoted above as the scribe of Add. MS. 31,055, a
collection of transcripts from the Recl and White Boohs of Hergest
and other sources.
Three Mid.-Tr. Hom.: " Three MddU-Irsh Homlies on the Lives of
Saints Patrick. Brigit, and Columba. Edited by Whitley Stokes.
(ne hundred copies privately printed.) Calcutta, 1877.''
Tit.: see " MS. Tit."

Toyal Troi (LL.): "Togail Troi. 'The Destructionof Troy', transcribed

from the Facsimile of theBooc o Leinster, and translated with a glos-

sarial index of the rarer words by Whitley Stokes, Calcutta, 1882."

Traeth.: see " Y Traeth."

Trc.: trcorois, i.e., in the Bretou dialect of the Pai/s de Trguier

(Breton Landreger), in the modern Dpt. of Ctes-du-N ord.
U: see"MS. ".
V: see "MS. V".
Vann.: vannetais, i e., in the Breton dialect of the Pays de Yannes.

The conventioual vannetais of books represents the dialect of the
neighbourhood of the town of Vannes (in Breton Gwsned). See
also " Bas Vann."
Vened.: Venedotlan, .e in the dialect of the district of Gwynedd, em-

bracing all N. Wales not included in " Powys", q. v.
Vesp.\ see " MS. Vesp."
Vis.: see " Her. Vis."

Vocab.: Vocabulary. See " Corn. Vocab.", " W. Lleyn".
vol.: volume.
W: see " MS. W".

AV.: (l)WalesorWelsh. See "N.W.", "S. W.", "O.-W." W.: (2)West.
W. (ilam.: Western Glamorganshire, i.e., all of the county lying west
of a line drawn from Merthyr Mawr (near Bridgend) to Aberdare.
(Camhr. Journ., iii, p. 244.)
W. Lleyn, Vocb.: The Vocabulary of the poet William Lleyn (or
Llyn; 1540-1587). The autographformsHenw MS. 122; L.Munis'
transcript in Add. MS. 15,055 has been used by Dr. Nettlau.
Williams: see " Caledfryn", "Thos. Williams", " Wms., Hgt. MSS."
Williams ab Ithel: see " Barddas, etc", " Dosp. Ed."
Wms., /////. MSS., ii: = the parts of Canon Robert Williams' Selectipns
from the 1 1< ■n;/irrt MSS., vol. ii, that are not taken from fche Llỳfr
tiiri/n Rhydderch, i.e., all from p. 284 to the end of Part V at
p. 34o. (MS. sources unknown, but can hardly be earlier fchan 15th
cent.) For the earlypart of fche vol., see "/./. Gw. /i'A."; Part VI,
corapleting the vol.,is unpublished.
Il'. Il'. /;. II'.: =The initials of fche late Mr. Wynne of Peniarth,
quoted under "MS. A, etc." (</. v.) as an authority for the dates of
some MS3. in his collection of Hengwrt MSS.


X: see "MS. A".
Y: see"MS. F".

Y Bed.: Y Bedyddiwr, Caerdydd [Cardiff], 1849, etc.

Y Cylchgr.: Y Cylchgrawn.

Y Cymiar.: " Y Cymmrodor; the Transactions [and Magazine] of the

Honourable Society of Cymmrodorion", Loudon, 1877, etc.

Y Cyf. Dyfyr (Ruthin): Y Cyfaill Dyfyr [i.e., Difyr].

Y Drych Chr.: Y Drych Christianogawl. Ed. by Rosier Sinith. 1585.
F God. (or God.): Y Gododin. See " B. of An."

Y Gweith.: Y Gioeithiwr, Aberdare.
F Gwrou Cymr.: Y Gwron Cymreig.

Y Gwyl: Y Gwyedydd. Bala, 1823-1837, 14 vols.

F S. Greal: " F Seint Greal .... Edited with a Translatiou aud [so-
called !] Glossary, by the Revd. Robert Williams" (London, 1876).

Note 1. This text was transcribed from Ifengiurt JIS. 49, a
MS. of about 1380-1390.

Note 2. The printed work comprises Parts -IIl (forming vol. i)
of the Editor's Selections from the Hengwrt MSS. See for the
remaining Parts, " Ll. Gw. Rh. 1 ' and " Wms., Hgt. MSS., ii".

F Traeth.: Y Traethodydd. Dimbych [= Denbigh], Treffynnon
[= Holywell], 1845, etc.

Yn.: see " Cyfrinach, etc."

YrAms.: Yr Amserau.

Yr Arw.: Yr Arweinydd, sef Newyddiadur wythnosol, Pwllheli, 1856-9.

Yst. de Car. Mag.: Ystorya de Carolo Magno, from the Red Book of
Hergest [transcribed by Mrs. John Rhs, and] edited by Thomas
Powell [sic], M.A. London; Printed for the Honourable Society
of Cymmrodorion, 1883." (Anotheredition of Campeu Charlymaen,
q, v. See also " of Herg. v )

Z: see "MS. Z v .

Z 2: The second edition of Zeuss' Grammatica Celtica.

Zeuss., G.C 2: ) Also " G. C?\ q. v.

Zeitschr.: Zeitschrift. See " Kuhn".

Note'. I have compiled the above rough Index to Dr. Nettlau's
abbreviations, etc. (which makes no pretensions to completeness of
detail), from such books and other authorities as I had access to here,
in order to make his learned article more generally useful to the readers
of F Cymmrodor. E. P.

Darowen, Cyfeiliog; Medi 30, 1888.



Adolygiad diweddaraf: 2011-02-15 23.48;

Ble'r wyf i? Yr ych chi'n ymwld ag un o dudalennau'r Gwefan "CYMRU-CATALONIA" (Cymreg)
On sc?
Esteu visitant una pgina de la Web "CYMRU-CATALONIA" (= Galles-Catalunya) (catal)
Where am I?
You are visiting a page from the "CYMRU-CATALONIA" (= Wales-Catalonia) Website (English)
Weə-r m ai? Yu a-r vziting ə peij frm dhə "CYMRU-CATALONIA" (= Weilz-Katəluniə) Wbsait (ngglish)

Archwiliwch y wefan hon
Adeiladwaith y wefan
Gwaith cynnal a chadw
Cysylltwch ni trwyr llyfr ymwelwyr

Free counter and web stats