4?

\AM)l

(5(4

s^^.

OLD-BRETON GLOSSES.

"^

EDITED By

WHITLEY STOKES,
CORRESPONDANT DE L'INSTITUT DE FRANCE.

CALCUTTA:

1879.

^^
\ S^

JlY^^/^^^/ ^. .

("^^^^^

T^^^^^p l^/--^Uv

^^^^^^^

^/i^

y^^
c^i^kL

^W^^ ^M

JJ^V^

^^.^\y-S^

6k^

J^oc^^y^

^JaU/a^ (i^oMiA

C'^A.^Ui^

A.

(11.

Jn..

Ux^//

^i^'.o/

^j^t^

^ -V

S-

'^AA-^dL

l^^^^^J^ 4y^(XA^

Au^xxhJ

f (^

OLD-BRETON GLOSSES.

1
CDITED BY

WHITLEY STOKES,
CORRESPONDANT DE L'iNSTITUT DE FRANCE.

{_Fifty copies

privately printed.l

CALCUTTA:

1879.

CALCUTTA
PRINTED BY THE SUPERINTENDENT OF GOVERNMENT PRINTING,
8,

HASTINGS STREET.

PREFACE.

Two
Welsh

sets of Old-Breton glosses are generally

known

one, those in the Bodleian Eutychius, printed as

Old

in pages 1052- 1054 of the second edition of the Granimatica Celtica ; the other, those in the Luxemburg fragments of the Hisperica Famina, printed, also as Old Welsh, in pages 1 063-1 065 of the same book, and, with corrections by Professor Rhys, in the Revue Celtique, tome I, pages 348-350. These glosses must be Welsh, Cornish, or Breton, That they are not Welsh, has, I think, been proved by Mr. Bradshaw, who points out, inter alia, that forms like doguo-misur., do-guo-hintiliat, do-uo-hinuom, do-guo-renniam, do-uo-louse, cannot be Welsh, in which language similar compounds begin with ^z-^?<o, dyguaz= dyo. That they are not Cornish is rendered probable by the occur-

rence in didanuud

(gl. elicio) of

the

compound

preposition di-

dan{=di-\-tan), for which

e\i\\e.r

in

Middle Cornish

we have always

yn-dan or a-than, and by the absence of the AngloSaxon signs for the sounds th and w. It follows, then, that they are Breton and these arguments are re-inforced by palaeographical considerations and by the general agreement of the forms occurring in the glosses above referred
;

to with those in the glosses

now

printed, which are incon-

testably Breton.

The
I.

mss. from which the following glosses are taken

:

are six in number, namely

The Berne codex

167,

containing scholia on Vergil


iv

Preface.
fifty-seven

and

Old-Bretoa glosses.

This
:

is

described, as

follows,

by Professor Hermann Hagen*

" Codicem Bernensem 167 (Q saec. Villi X...eadem sed non omnia codicis B scholia continere iam erat notum...Contini" cum figura astronomica entur autem eo codicc.haec fol. I/ quaedam ex Isidori de rerum natura libro excerpta f. I" III" titulo INCIPIT ARGUMENTUM IN VIR. de Vergilii uita scrip:

tisque nonnulla medii excogitamenta...,

f.

Ill"

INCIPIT CAR-

MEN OCTAVIANICAESARISDE VIRGILIO...Manurecentiore

Richardus,' lectoris scilicet... f. IIIF ^V" subscriptum nomen INCIPT. EXPST. SERVII. GRAM. IN BVCOLIC. IN LIBRIS
'

GEORG. ATQUE AENEIDUM

Aen,..Secuntur
f.

Seruii praef. in Bucol.

Georg.

VITA VERGILII POETAE f. VP decern uersus. ..Inde a fol. VI." fol. XX^ Vergilii Bucolica, INCIPIT DRAMATICON MICTON. MELIB. TITVR., cum scholiis Bernensibus, f. XX^ XXXIII Georgica cum schol. Bern. Sequitur inde usque ad finem Vergilii
V^ uersus
uitag

interpolatae...Sequitur

f.

Codicem in Britannia uel Scotia scriptum esse testanbaud paucae Iro-Celticae, quas ad uirorum doctorum, quibus talia sunt curae, usum sub uno omnes conspectu iuvat
Aeneis
tur glossae

proponere."

Mistakes
daliu,
'
'

such

as

'

frigora
,i.

.i.

guascotou,'
'

'fuscus
'

.i. .i.

tribuli

spine labet

gloiatou

and

obnixus

utgurthconeW seem to shew that the glosses in this codex were transcribed from another ms. by an ignorant or
careless copyist.
2.

copy

of

Amalarius

De

divinis

officiis,

written (ac-

cording to Mr. Bradshaw) A.D. 952, apparently at Landevennech. It afterwards passed over to Canterbury and is now at Corpus Christi College, Cambridge, No. 192.
of the Collatio Canonum, a 3, 4, 5, 6. Four copies work apparently compiled in Brittany at the beginning of These mss., says Mr. Bradshaw, range the eighth century.

from the ninth, or ninth-tenth, to the eleventh century. All four have Old-Breton glosses and must have passed out of
Brittany during the
to

Corbey

in

Picardy,

Norman now

desolation of the country,

in

one

Paris; one to Glastonbury,

* At pp. 690, 691 of his book entitled Scholia Bernensia ad Vergilii Bucolica atque Georgica, edidit emendauit praefatus est Hermannus Hagen.

Commentatio ex supplementis annalium philologicorum seorsum

exscripta.
Lipsiae, in aedibus B. G. Teubneri,

typis

MDCCCLXVII.

Preface.

v
British

now at Oxford one to Canterbury, now in the Museum and one to Fecamp in Normandy, now
;

also

in

Paris.

The
Lat.
1

first

copy,
is,

2021)

now in the Bibhotheque Nationale (Ms. according to Mr. Bradshaw, written by a

Abbot Haelhucar. some extracts 'de

man named Arbedoc by

It

leave of his contains three Irish phrases * among

It

disputatione hibernensis senodi.'

has also the canons of

Irish n as r) as canones addamnari uel addominari.' At the top of one of the pages of the second copy occurs the address matguoret benedic mihi,' a name which seems to point to Brittany, while one oi the probationes pennae in the margin serves, in Mr. Bradshaw's opinion, to connect it with the Luxemburg fragments of the Hisperica Famina. The third copy, which has been greatly injured by the fire of 1 731, is the Cotton Ms. Otho E. xiii. The scribe glosses 'quidam patricius by tiranus,' which reminds one of the tiranni or machtierns of the Redon cartulary. The fourth copy (Bibliotheque Nationale, Ms. Lat. 3182) is, says Mr. Bradshaw, the latest of all the mss., but even this cannot be later than iioo. For the fifty-seven Berne glosses I am indebted to Pro' '

Adamnan, described (from misreading the

'

'

fessor Sophus Bugge of Christiania. He accompanied th^m by a commentary in German, and, except as regards Nos. 3,
5,

II, 12, 23,

28,

29, 33, 37, 38, 44,

little

47, 48, 53,

his.

on these glosses are

afterwards sent

but a translation of

my notes He also

me a new collation of the Berne glosses made by Professor Hagen, which furnishes one new gloss (uentrem .i. tar) and corrections of six of the others as printed in the

* " Hec est poena magi uel uotiui mali si credulus id demergach uel preconis uel cohabitatoris uel heretici." Here id' stands for id est and demergach is perhaps to be equated with dilergach in aith-dibergaig renegades' ? Fis Ad.
'

'

',

'

A man
throscho." fasting.

has to fast on "panibus qui efficiuntur de tertia parte coaid stir Here we should read coid sirthrosctho of the food of a long
'

man has to live as a commutation for a year's penance so much time fordohorfiit' besides having to say so many extra psalms a day. This ^\io\AA\)Q for doborphit on water diet:' 'in pane et aqua' is the usual
' '

phrase.

vi

Preface.

Scholia Bernensia.

The

rest of the glosses

owe
in

to the kindness of

Mr. Bradshaw,

upon them.

January 1877, with a suggestion In so doing I have been greatly aided by the remarks contained in a letter from Professor Bugge
dated the i6th September 1879.

now printed I who sent them to me, that I should comment

W.

S.

Simla,
21st October i8yg.

OLD-BRETON GLOSSES.
-THE BERNE
.

I.

GLOSSES.
.

18 ligustra .i. melgabr 30 47 uiolas .i. uileou. Ill 55 quandoquidem .i. annaor IV 23 conabula .i. mabcauuelou 47 fusis .i. aguirtitou. VII 42 Rusco .i. ethin VIII 34 supercilmm .i. guorail .X. 19
II. 8

E.

frigora id est guascotou

elestr
.

hibiscum
.

.i.

opilio

.i.

ousor

38 fuscus
.i.

.i.

dah'u

41 Serta
.i.

.i.

couarcou.
.

G.
cauell

44putris
tilia

hue. 153 labet

limncollin
.

gloiatou
X.

166 uannus

.i.

.ly^
.i.

.i.

178
.i.

cylindro
.

acronnmain
.i.

201

lembum
c/
.

cauhal

262 obtusi

/rwf^

266

fiscina

ser uel

308 Auritos

.i.

scoharnocion

foedam .i. daureth 363 fulice .i. 388 comix .i. cornigl 392 putres .i.
. .

309 Stupea .i. iscartholion .323 guilannou 364 ardea .i. cora'd
.
.

bocion.

G. II 381 proscenia
lances
.i.

.i.

racloriou
.i.

discou .413 rusci

.i.

ethin

389 Oscilla .i, luscou 449 tiliae .i. limncollou.

.

394
.

G. Ill 100 notabis

148 musca uolitans cosmid 564 pabule
.

agnosces uel signabis

.i.

.i.

ercenthidite
.i.

.i.
.i.

attanoc
huital.

clehurin

406 serum

?neid uel

G. nil 120 Intiba .i. cocitou .122 uentrem .i. tar. 131 uerbenas .\.ueruencou 168 Fucos .i. satron nelguohi 388 ceruleus .i. duglas.
. .

Aen.

726 laquearibus

.i.

attinou uel acepriou.

.

Aen. II 29 manus .i. bodin 85 cassum .i, ihepcorim. i8opatrias broolion 236 canabina .i. coarcholion 236 lapsus .i. libiriou uel stloitprenou 646 iactura pritiri. 22 tumulus .i. fcfA. Aen. Ill 20 auspicibus .i. doromantorion
.i.
. .
. .

158 idem uenturos 31 lentum limn 92 cortina .i. esceilenn nionuret 289 transtris .i. libiriou 549 antemnarum deleiou.
.

.i.

Aen. nil 131 plagae

obnixus
.i.

.i.

guinodroitou

MenzhnXdi guinuclou

332

utgurthconeti.

II. GLOSSES IN

AMALARIUS.
192.
.

Corpus Christi College, No.

uiczx'wxs,

amsauath
sic

segnitia blinder
[in

]\2ih\\is

camadas nihilominus
.

nihil

minus

quoque

marg.]

.i.

non minus

nahulei uel int

coucant.

. .

Old-Bretoji Glosses.

III. GLOSSES IN

THE COLLATJ,0 CANONUM,

I

ist Copy.

BiBLIOTHEQUE NaTIONALE, Ms. LaT.

zxc^\Vcnferafib
.

202 I.

inopportunius seingerit ejitcrafho hciianemdoguot

.

turpi dour et it angruit

aeditui id costadalt
.1.

accommodata dehlouetic.
.

fastu amuoet

piacula
.1.

abscenia
.

.1.

caul

istriones
.

.i.

guanorion

phitonistarum
art hui.

torleherieti

agipam gulcet

uos fascinauit ar uiio

IV. GLOSSES

IN.

THE COLLATIO CANONUM,

HaTTON,
42.

2nd Copy.

BiBL. BODL. Ms.

ultro
.1.

aiul

ultro
.

ambit uel inopportunius se ingerit aruanta

.

incaenis inuanetou
clou
.

antropas (leg. andronas) dadluo

acitamenta

fenus

.i.

endlim.

v.-GLOSSES IN THE COLLATIO CANONUM,

Mus. Britt. Cotton Ms. Otho E. XIII.
andronas
climatibus
.i.

3RD Copy.

dadlou
.

scurilis.i.

guaan nepta
.

.i.

nith
.

gomor mod
.i.

Rannou uorticem mentis

.

.i.

cunru7it
.

passae

ad7net

controuersiam controliaht
aaltin
.

agipam
.

.i.

latic

defer .i.^?</r/c. ferula

.

.i.

hirsutis aceruission
.

beluina rabies cmtnaret hoestol

curiae

dadlou

strutionem
::

.i.

Irof

noctuam

.i.

couafin

larum

.i.

trad

attacus deuo

epimachus hiunrun.

VI. GLOSSES IN

THE COLLATIO CANONUM,

3 1 82.

4th Copy.

BiBLIOTHEQUE NaTIONALE, Ms. LaT.

graciles salirocion
.

probum guohethe

bouello buorih.

Old-Breton Glosses.

COMMENTARY.
1.

guascotoii.

I.THE BERNE GLOSSES. The context is

et frigora captant.

Nunc
guascotou
frigora.
It
is

etiam pecudes umbras,

the

must gloss umbras, not

(as the glossographer says)

plural of gtiascot,
'shelter,'

now gwasked

'

abri,'

W.

givascawt,
'

now gzuasgod,
'

O. iT./oscadZ." 1028, now/bsgadk

shadow,' a lunar

shelter.'
lit.

In the modern Bret, gwaskaden

'

war

al loar

shadow over the moon, the meaning of umbra is retained. From gua-, by regressive assimilation for guo-, It./o-, tto, and scof, I\Iid.-Br. squeul, Corn, scod, O. Ir. scalA. Compare vmo-xtog as Mod.-W. cysgod is analogous to
'

eclipse,'

'

(r6(Txiog.

that scol

'

The plural ending -ou seems here to be original, for was an -stem is probable from Goth, skadu-s. In the Berne scholia to Eel. II 18 2. melgahr (gl. ligustra). Mel is therefore ligustra' is explained by 'flores papauerum.'

Corn, mill
hionou

(gl.

papauer)

(gl. uiolas),

Compare also O. W. mellCorn, mel-hyonen vigila.' As to gahr it must

1076.
'

Z?

be the Mid. Br. gaffr 'capra,' Corn, gauar, W. ga/r, O. Ir. habeo gabor, Gaulish gabro-, which is cognate with O. Ir. gaibiu papaver as caper is cognate with capio. Mel-gahr thus means caprae,' and is an improper compound like Penn-ohen 'caput

=
'

'

boum'
boum,'
3.
' *

2.?

126,
'

W.

costad-ali infra

No. 75, rj^^yt^^w

vadum
tlesir

Oxford,
(gl.
'

elestr

Penn-ichenn Z.^ 125, ial-cen, etc. hibiscum 'wild parsnip ?' 'mallow.?')
'

now

iris,' ^/^i-/r^7z

gladiolus

Cath., Corn, strail elester (gl. matta)

'flag,' 'fleur-de-lis,'
Ir.

Z?
'

1079, ^lestren (gl. carex) Z.^ 1076, V^. elestr

Ir. elestar , f-elestar , s-oilestar (gl.

iris,'

gladiolum)

Gl.

No. 795.

Probably cognate
alestrare, alistrare
4.
i.

with
e.

aXio-fj-a.

'

waterplantain.'

Ducange has

humectare.

uileou (gl. uiolas), a loanword.

5. afl(?r (gl. quandoquidem). Prof. Bugge compares the Corn, a r R. 882, and Mod. Ir. anuair 'when' this is the Mid. Ir.
:

inuair (Three Fragments,

122), the O.

Ir.

inn-uair (Issin tra in

cochoU
for or

roiarr

^ed for a
is

gilla innuair ut,

Breton ann for enn

the definite article,

borrowed from hora, just as borrowed from, or cognate with, culpa.

'

LL. 2i4.b.i).* The Old and aor (W. awr') is caul, infra No. 78, is for cut,

* Now that is the cowl which Aed asked of his gillie then.' The corresponding Welsh adverb \s yr awr a honno ae gweircheidw wy yr awr y kandeiryocont and that one guards them when they rave,' Y Seint Great,
:
'

ed. R. Williams, p. 418.

4
6.
'

Old-Breton Glosses.
7fl3caMf/tf (gl.
',

conabulai.e, cunabula).
mace,

A compound of ;wa(5
W.
cawell

child

O.

W. map,

Ir.

and cauuelou,
15, cauell
'

better cauuellou, pi. of

ca?<^//

(gl.

uannus) infra No.

'

'

berseau a gesir enfants/

All borrowed

cauell da quempret pesquet

'

nassa,

Cath., Corn, cawal,

'

corbis,' Z.2 819, A. S. cawl, ceawl

basket.'

from
ac),

the

Romance
is

cauuella,

which occurs

in the Cassel glosses.

is

7.

aguirtitou (gl. fusis).

Here a

the preposition

(W.

which

used

to express the ablative, as in

a cronnmain, a ninou,

a cepriou, a ceruisston, and perhaps occurs in composition in enter-a-fib No. 63, 2SiA guirtitou is the pi. of guertit, no^ gwerzid. Corn, gurhthit,

W. gwerthyd

'

spindle/

'

axis,'

Ir.

fersaid.

All
also

apparently, from a groundform *verttati, root VERT,

whence

Ch. Slav, vreteno and Mod. High German wirtel. 8. ethin (gl. rusco). So ethin (gl. rusci) infra No. 31. This

is

Com. eythinen
aittenn.
'

(gl.

ramnus) Z? 1077, W.

eithin 'furze,'

'

gorse,' Ir.
ostinv,

Prof.

Bugge compares,
supercilium).

doubtfully,

Ch.

Slav,

stachel,' Lith. aksztinas.

9.

guorail

(gl.

ail

O. W.
this ail

From

From guor super,' 2,? 905, and Kuhn's Beitr. VII, 390, 398, now ael. the Old Breton name Gurm-hailon, Cart. Red. 224,
'

ail 'brow,'

225, seems derived.

=Mid. W. heussawr pastor,' now heusor, Compare Mod. W. heus-leuen sheeplouse,' heuso 'to proThe stemword, according to Prof. Bugge, seems tect,' 'to shield.' Lat. hapsum vellus lanae' Gl. Isid. cited by Diez, Etym. Worterb. 4" Ausg. s, 512, s, v. aus, which in the Mod. Proven9al means For the change of meaning Prof. fleece,' O. Fr. heus de mouton. Bugge compares Skr. mesha fleece' and 'sheep,'
10. ousor (gl. opilio)
'

Z.2 830.

'

'

'

'

is an obvious mistake or duliu, du =^lr. </3 'black' (cf. du-glas, gl. ceraleus So liou (gl. neuum) Lux., Corn. infra No. 41) and liu 'color.'

11. daliu

(gl.

fuscus).

This

compound

of

Bret, disliu (gl. discolor),

W.
;

disliw, Corn, unliu

(gl.

unicolor)

W.

liu (gl. gratia)

Lat. livor

and

cf.

ApoUini Livio, Orelli 2021,

cou- for cov-,
initial p, geflochtenes

Livius,

Z?

109.
(gl.

12. couarcou

serta)

seems compounded of
Skr.

com-, Z^ 902, and arcou pi. of arc, which I take to have lost

and

to

be cognate with
hue

praqna

'

geflecht,'

'

korb,'
13.

BR.

ttAsxcu, plica, plecto.

(gl. putris), pi. bocion (gl.
.i.

putres) infra No.

is

27.

The

ms. has putris

buc

.i.

mollis.

This

now

written houc

'mou,'

'tendre,' 'dlicat.'

(gl. mollior), (57f-glas,

whence bogach

'

the Mid. Ir. (5w (gl. tener), compar. buigi Tain Bd Fraich, Mod. Ir. bog 'soft,' 'moist,' marsh and the English loanword bog. If, as is
It is
'

Old-Breton Glosses.
possible, the b here
cian,
is

5
cweccan, cwa-

from^z',
c

we might compare AS.

though the Teutonic

does not correspond with the Celtic.

in the margin, opposite Georg. I

14. gloiatou.

The ms. has

.i.

153, TribuU spine labet

gloiatou.

The

context

is

Lappaeque

tribulique interque nitentia culta

Infelix lolium.

'

Here the copyist has obviously blundered, for lappae znd gloiatou can only refer to nitentia.'
'

'

labet

'

should be

'

gloiu,

now gloyw, gloew,

Ix.

'

limpidus,'

'

lucidus,'

Compare O. W. Kuhn's Beitr. IV

'

411, O.
Z.2 842.

gl,

Z?

105.
i.

For the

suffix cf. ung-coffat

co-uterinus/

cleizyat 'mantinus'

e.

scaeva, -^//a/' hilaris,'^?ra/' lamentabilis,'

15. cofinus uel cauell (gl.

uannus).

See above. No.

6,

16. limncollin (gl. tilia), pi. limncollou (gl. tiliae), infra

No. 32.

These

are

compounds
Ir.

oli??in (gl.

lentum), infra No. 50 =r

'

W.

lly/n
'

'smooth,'

slemon,

{slemna

gl. levia 2,? 'j'j()),slemain

lubricus

(ex *slibna, * slibni) and

'hazel,' Z.2 791,
17.

coll-in, coll

(ex *cosla)

W.

collen, coll

Kuhn's
(gl.

Beitr.

a cronnmain

cylindro).

VII 396. Here a

is

the preposition

already noticed
'

(No.

7),

and cronnmain, a compound of cronn

round,' O. Br. cron (gl. tornatili) Lux.,

Beitr.

now

krenn, O.

W.

crunn,

Kuhn's

VII 391, O. Ir. cruind,2in. main 'stone,' now maen, Corn. 7tien, W. maen, Kuhn's Beitr. IV 404. 18. caubal (gl. lembum). Borrowed from the Latin caupulus, like W. ceubal 'ferry-boat,' 'skiff,' Old-Northumbrian cuople
and Scotch
coble 'a small fishing-boat.'
19.

'navicula,'

truck

(gl. obtusi).

Borrowed from the Latin truncus,

like

Mid.

W.
is

truck,

Kuhn's

Beitr.

IV

423,

trouck

used only as a substantive,

The Mod. Br. meaning coupe.' The Corn.

now

trwck.
'

ireck glosses truncus, Z.~ 1077,

20. ser uel cest (gl. fiscina,
'

a small basket of wickerwork,'

'

measure
ioxper
other
'

for milk
'jar.'
'

').

Here
'

I think ser

must be borrowed from the

Ir. coire

Latin jerza

Possibly, however,

it is

bassin

'

chaudron
'

=
It

W.

pair,
'

a mistake of the scribe ex * kvapria. The

'

word cest, now kest, corbeille,' panier,' is, like W. cest a narrow-mouthed basket,' borrowed from Lat. cista. There is a

W.

cist

according to Rhys.
(gl.

occurs in

Y Seint

Greal 99.

21. scobarnocion

auritos), pi. of scobarnoc 'auritus'

skouarnec 'hare,' Corn, scouarnoc (gl.lepus),

'

= now W. ysgy/arnog, and

(if it be genuine) O'Reilly's sciberneog. The stemword is scovarn. Mid. Br. scouarn ear,' now skoarn, (Corn, scouarn (gl. auris),
'

Mid. W. eskeuarn, now ysgy/arn. These words have nothing to do (as Prof. Windisch, Kuhn's Zeitschrift
scovern,
scoforn.

Old-Breton Glosses.
429, supposes) with
the root sku
'

XXI
xoa,

schauen'

or the

Gr.

KOUSU

They
50,
I
:

are unquestionably, as Prof.

Bugge

suggests,

loans from the Latin caverna. For the meaning he compares Pliny

N. H.

XI

'

cavernas

habere

aurium

loco.'
'

The

s is

prosthetic, as in sclacc

= Fr.
'

'glace,' sclaer
'

he compares Port, covo

pares Bret,

hollow,'

Schuchardt's Vocalismus I
22. iscarthohon

178.

For the cava hole,' and other words in For the d oi scobarnoc he comFr.
clair.'

beni'n, hilen, bisalg, beronic,

rambre.
of

(gl.

stupea).

Plural
(infra
'

an
'

adj.

iscarthol,

formed
stantive

like broolion, coarcholion

Nos. 45, 46).

Z.^

The
Ir.

sub-

iscarth

is

W.
i.

ysgarth
e.

offscouring,'

excretion,'

O.

Ir. escar/ (gl.

peripsema,

Trep/^/rj/xa)

800, Mid.

escart

(gl.

scupa, leg. stuppa),

ys-,

Z?
'

904,

Ir. es'

Ir. Gl. No. 254. From the prefix is, W. es-, and earth now Jcarz raclure,' ordures,' W.
'

'

earth

hemp,'

tow,'

'

oakum.'

The context is " Et foedam glomerant 23. daureth (gl.foedam). tempestatem imbribus atris." This is obviously cognate with douretit infra No. 64 and possibly with the O. W. ddfraud in the
phrase defraud atuis Juvencus, pp. 2, 18, Kuhns Beitr. IV 390, VII 412, where atuis is perhaps the O. Ir. athiss 'opprobrium/
1 conjecture that
it

stands for *draud, dravja, a derivative from

also the

the root dru,

whence
368.

Welsh drewi
e.

'

olere,'

'

foetere,'

Rhys,

Rev. Ceh.
24.

i.

gu I'lannou
(gl.

(g\. inlice
'

i.

fulicae 'coots'), pi. of guilann

Mid. Br. goelami

giiilan

meaning
gwylo.

to

now gwehm, goelan, goilan, Corn. alcedo) = Ir. foilenji, Z? 778. All from a root wail, M. Br. goelaff, now gwela, W. gwylaw now
ulula
'

Cath.,

The words under

notice have furnished two of the rare

Celtic loanwords in English

and French, namely, Eng. gull and

Cdiih..,

Fr. goeland.
25. corcid (gl. ardea).

Mid. Br. quercheiz,

now

kere'heiz.

e, y in the forms just cited must have and Professor Bugge may be right in comparing the Gr. xQpKopa, Fick^ 141. The O. H. G. hreigir, N. H. G. reiher A. S, hrdgra, O. N. hegri (for hreigri) may possibly be cognate.

Corn, eherhit i. e. kerhtth, mistake for cerchid, the

0,

W.

eryehydd.

If eorcid is not a scribe's

sprung from

26. cornigl

(gl.

cornix).

loan from Lat. eornicula.

articulus,
13.

One

would have expected

27.

corniel as artiel

bocion (gl. putres).

(gl.

from See above. No.

pi.

Z}

817.

28. racloriou

proscenia),
'

of

raclaur 'proscenium,' a

compound
677,
Z.2 1054,

of the prep, rac

before,'

Corn, rak, rag,

W.

rac,

Z?

678, 679,

now rhag

Mid. Br. laur,

and laur (gl. solum) Buhez Mabden, 280, kur aere,' 'aera'
(Skr. prk, Ebel),
'

Old-Breton Glosses.
Cath.,

j
vel

now

ler

'

sol,' 'aire,'

Corn,

lor (gl.

pavimentum

solum),

W.
and

llawr, It. lr.

are identical with A. S. Jlr,

These words have all probably lost initial p, Eng. ^oor, N. H. G. ^ur.

29. luscou (gl. oscilla).

The

context

is

tibique Oscilla ex alta suspendunt

moUia pinu. (G.

2,

389),

and here, of course, oscilla is the pi. of oscillum, the little masks of Bacchus hung up in vineyards. But the glossographer obviously took it to be the pi. of oscillum a swing ;' for luscou is the pi. of lusc, which is cognate with the modern liiska osciller,' luskelladiir 'oscillation,' Mid. Br. quef-lusqui remuer,' tressaillir Pohnes Bretons, 204. The Corn. Icsk a cradle,' Lhuyd, Arch. Brit. 53, s.
' ' *
' '
'

V,
'

cunabula, 69
' '

s.

v.

incunabula,

is

perhaps cognate.

The

Irish

and letnv-lusca,' which Lhuyd also quotes, are to me unknown. But the Ir. luascad a rocking,' luascach waving,' luascan
luska
' '

'

cradle

'

may be

cognate.
Littre

Chevallet connects the Fr. tocher with

prefers Diez's etymology from

these words.
liicke
'

But
(gl.

MHG.
discus.

loose.'

30. discou

lances),

pi.

of

disc,

borrowed

from

Hence also Ir. iesc (gl. lanx) Sg. 20% AS. disc {dix), Engl. dish. From the diminutive <f/jaJ?itj comes W. disci (gl. lance) Juvencus,
p. 59,

Mid.

W.

dysgvl,

Seint Greal 144.

31. ethin (gl.

rusci).

The ms. has Exiguum

'

rus

rusci

id.

inculti agri ruse ethin.

See above, No.

8.

32. limncollou (gl.

tiliae).

See above. No.

.i.

16.

33. ercentbidite (gl. notabis

is

agnosces uel signabis).

cent another prefix
'

Here

er-

the particle found in em-er-hedaf, but generally written ar, Corn.

,

ar-

er-

yr,

W.
and

ar- Z*
didite is
te

900

from cant Z.^

Z.^ 556,

901

= KciTOL;

the Corn, bythyth

eris,'

hydy, with suffixed

of the pron. of the

is

2d

sg. Z.- 370, 507.

W. The

corresponding verb in Welsh

arganfot Z.^ 574, 907, arganuot,

'to behold,'
ctbuith,
'

Y
'

Seint Greal 272,

now arganfod

to discover,' 'to
ctbuid
'

perceive.'

The
\.

Irish

cognate seems
(gl.

sensus,' Z.^ 992,

= W. canfod.
clehurin

sentire,'

34. attanoc
for atanoc, pi.

musca

uolitans).

atanocion (gl. alligeris) in the

Here attanoc is Luxemburg glosses.

'

The cognate Welsh

derived from atati
ein

adjectives are adnog, adeiniog

'

winged,' both

penna,'

pi.

ataned,

an older form of which,

(=Ir.
6''.

e),is preserved in the gloss etn-coilhaam (gl. aspicio)

Eutych.

These words have

all lost initial

p and

are connected

with TTSTo/xaj and other words treated by Curtius G. E. No. 214.

clehurin
35.
is

the Mid.
\xq\

W.

cleheren

'

tabanus,'

now clyryn, cleren

;//(/ is

'

fly.'

meid

cosniid {q\.

serum).

Here

Corn,

nieilk

Old-Breton Glosses.
'

cos-mid

(Lhuyd, Arch. Br. 149,^ 289,*^ W. maidd, Ir. medg whey/ and whey-curds :' cos, now kaouz, W. caws is W. caws maidd * cheese,' cosyn a cheese/ Ir. caise, are all borrowed from the
'

'

Latin cseus.
36. huiial uel uerrucae (gl. pabule
/.

e.

papulae), the collective

to *huitalenn,

c'houeda,

W. chwydalen 'blister.' The Mid. Br. and W. chwydu 'to vomit,' are cognate.
Lat. vomica
intiba).
is

huedaff

now

So, as Prof.

Bugge remarks,
37. cocitou

cognate with vomo.

difference

(gl.

Notwithstanding the
the pi. of

of

meaning between
(gl.

cicuta

and intyhum,
is

this should, I think,

cecit,

be

cecitou

intyba), where cecitou

now

kegid, borrowed,

like

W.

cegid,

Corn, kegaz, from the Latin cicuta.

sint,
:

The

scholia

Bernensia have " Intuba quod intus cava

stalks of the cicuta or

quasi tuba."

The

hemlock are hollow see Lucr. 5, 1382, Verg. E. 2, 36. Professor Bugge prefers to connect cocitou with W. cecys plants with hollow stalks,' [whence Eng. kex .?] ceccysen canna,' Davies, and ceg throat' = Ir. scig. But this last seems the Br. chouc, 'on lit scouc dans un ancien ms.' (Le Pelletier).* This gloss stands over the word nee' 38. tar (gl. uentrem). in the line Cresceret in uentrem cucumis, nee sera comantem. It must refer to uentrem and is the Ir. tarr, Mid. Br. torr. Corn, and W. tor: it reoccurs, spelt tor, infra, No. 71. A loanword from Lat. verbenaca. 39. ueruencou (gl. uerbenas). The c is perhaps to be compared with the g of the Corn, tivul-g-ou,
'
' ' ' '

'

Z2

847.

40. satron Mt\

guohi

(gl.

fucos).

Here satron
'
'

is

the

collective
'

of the singulative satronenn.

'

Mid. Br. sardonenn bourdon,'

assillus,'

fuscus

'

(leg. fucus) Cath.,

(gl.

now

sardonen

frelon,'

'

taon.'

The

Cornish sudronenn
guohi.
(gl. vespa).

The The

fucus) should apparently be sadronenn. singulative of this is * guohienn Corn, guhien

Lat. vespa, Ch. Slav, vosa, O. Pruss. wobse, Lith.

:

vapsa seem cognate

cf.

uher in Corn, gurth-uher

is,

=
'

vesper.

41. duglas (gl. ceruleus)

like

W.

aulas, Ir. dubglass, a

com'

dub (= *dugva, dunkel) and glas viridis (gl. glaucum) Lux. cognate with glastum, Kuhn's Beitr. IV 398, VII 389. The context (Aen. I 42. aninou uel acepriou (gl. laquearibus).

pound

of du, Ir.

726)

is

dependent lychni laquearibus aureis

Incensi.

As to the Ninou is the

heaven.'

prep,
pi.

a before

the ablative,

see
'

No.
'

7,

supra.
'

of nin

=
(gl.

W.

nen

'

ceiling,'

vault,'

roof,'
'

the

Corn, nen-bren

laquear), Bret, nein 'sommet,'

s. v.

faite,'

* Pomes Bretons, p. 181,

chouc.

'

Old-Breton Glosses.
*

There is an Irish nion (leg. niti), which O'Reilly Heaven, the expanse or firmament ', and which he exemplifies by Pattraicc friheasgytamh Ninne (leg. nine) Patrick when ascending to heaven.' This may be the Irish cognate.
me,'
'

comble.'
'

explains by

'

'

'

cepriou

cepriou gl. tignae,

Luxemburg

Gl.)

is

the pi. of cepr.

Mid. Br. quepr 'chevron', now kebrra.., Corn, keber (gl. tignum), W. cebr, from Med. Lat. caprio, a derivative from caper, as Fr. chevron

caprionem.

43. bodin (gl.

manus).

The

context (Aen. II 29)

is

Hie Dolopum manus, hie saevus tendebat Achilles, Compare bodin (gl. phalangem) Lux. pi. bodi7iiou (gl. phalanges), Mid. W. bydin, Z? 90, Ir. buiden, Kuhn's Beitr. II 1 74. Root bhadh.
44. ihepcorim (gl. cassum).

The

context (Aen. II 85)

lugent.

is

nunc cassum lumine

Here i, for in, is the preposition, later en, and hepcorim is an abstract substantive, formed like diprim Z. 821, molim, etc. and = Mod. W. hebgori to dispense with,' to put aside,'
' ' '

to omit.'

It

is

compounded
'

of

hep,

the preposition

meaning

'

without,' Z.*

and corim
'

679, W. heb, Ir. sech, Lat. secus, compar. sequius, Ir. cor to put,' ' to cast.'

45. broolion (gl. patrias), pi. of brool, a deriv.

land,'
'

from bro
Ir.

brogae Galli

agrum

dicunt' Z. 207

O.

mrug

= W. hro = O. N.
.i.

mork, Zend merezu and cognate also with Lat.

margo and Goth.

cana-

marka.
46. coarcholion (gl. canabina).

The ms.
kiier,

has uincula

bina

.i.

coarcholion.

This

is

the pi. of coarchol, a derivative

from
flax.'

coarch

now koarc'h,

kouarc'h. Corn,

47. libiriou uel siloitprenou.

Prof.

W. cywarch Hagen says

'hemp,'

'

that this gloss

is on the margin v. 236 of the second book of the Aeneid over the words " Lapsus sine rotunda ligna quae rotis subponuntur."

The

context

is

235. Accingunt

omnes

operi,

pedibusque rotarum

Subjiciunt lapsus.
'

They put

the runnings of wheels under

its

feet

and

to the rollers over

which Sinon's horse was dragged into Troy

the glossographer

doubtless refers.
is

As

to

libiriou,

pi.

of

libir, I

think the second /

an

irrational
llyfr,
is

vowel and that

libir

should be

equated with the Mod.

car
'

W.

that part of a drag

which

which occurs in the phrase lly/r on the ground.' It is just possible

with this
lly/r.

that Gr. eXt^pos,

O.H.G.
slide,

In stloitprenou,
slidan,

s/e_far may be connected we have a compound of stloit

for sloid (A.S.

Eng.

to

O. Norse

sledht

'

sc\i\ii\.Qn')

and prennou pi
Ir.

Fick^ III 359 of prenn. Corn, pren

(gl.

lignum),

crann, Lat.

lo
quernus.
slodyys,

Old-Breton Glosses.
Our
O.
the
stloit,

like the Ir. slaod

'

sledge,'

W.
'

ysled,

Corn.
stloii
stlejel,

2318,

appears

to

be

a loan-word.

From

descend

stlejuz, the

modern stlej 'rampement/ stleja ramper/ final / becoming the lingual sonant spirant,
2in.

as

in

rouejou infra No. 56, egetm, nigal,

pmigejt.
'

s-t-laon

For the intercalation of / between s and /, cf. s-t-labez anguille and perhaps es-i-lam tonnement.'
' '
'

ordure,'

48. pritiri{g\. iactura).

'

The
on

context

is

'

facilis iactura sepulcri,'

the loss of a

tomb

will fall
.1.

me

lightly

;'

over iactura
'.

is

written

'

dampnum

vel pritiri

proiectio

mea

in sepulchro

The

glosso-

grapher seems to have regarded iactura as possibly meaning 'consideration' (cf. iaciantem pectore curas, Aen. I, 227). For
pritiri
is

the
'

modern
anxiety,'

pridiri
'

'

soin,

'

'

souci,'
cf.

'

sollicitude/

Mod.
(gl,

W.

pryderi

deep thought
auspicibus

:'

O.

W.

preteram

perpendo).
49. doroma^itorion (gl.
.i.

considerantibus).

do
do,

is

the Bret,

and Corn, form

pi.

of the preposition of *romantor,

= W.
'

Here
du,

di, Ir.

and romantorion the

equates with the Lat. praemonitor.

which Prof. Bugge There (W. rhy-, La.t. pro) here

corresponds with Lat. prae-, as in

'

W.

rhy-farnu

to prejudge,'

/JWE'-judicare.'

50. cjioch (gl. tumulus).

This

is
'

the Mid.-Br. knech, quenech,

'

quenechen7t,

now krach
'

'

monte,'

cnwc
'

'

hump,'

hillock,' Ir. cnocc.

'

tioned

elevatio,'

word with W. cwnwg cwn summitas,' altitudo,' and

' '

petite montagne,' W. Zeuss connects the last mensummitas, culmen,' erchyniad

tertre,'

the

names Cuno-bilinus,
(oprj)^

Cuno-tamus,
cynia
silva,

Cuno-maglus,

Maglo-cunus, 'Ap-xovia
vii.

Her-

Her-cuniates, G. C.^ praef.

92,101.

51. limn (gl. lentum).

52. esceilenn
(gl.

cortina).

See above, Nos. 16, 32. Here, as in the Old Welsh gloss
vii

lenn

(gl.

cortina)

Mart. Cap. Kuhn's Beitr.

409, cortina
'

is

understood to be a 'curtain' ('velum expelle*), and esceil-enn is a shade,' phansingulative form, cognate with the Mid. Ir. scail
'

tom and
'

53.
.

modern Gaelic scail veil,' curtain.' The context is nionuret. idem uenturos (.i. nos) tollemus in astra
the
'

'

nepotes.

'We
come
pi.

it

is

that will raise to the stars the descendants that shall

after you.'
is
'

Here ni
2r.

we,'

Z?

369, on

is

the possessive pron. of the

first

person

an abstract noun, a mistake for, or a corruption of, unet, W. U7iyd, meaning, as a substanunitas,' or, as an adverb, as one,' like,' the same as.' Comtive, pare the Mod. Bret, phrase ni hen unan nous memes.' An384,
' ' '
'

and uret may be

'

Old- Bret on Glosses.

other conjecture
is

that

tirel is

derived from ur

= the

Irish infixed

n in anlaut (Uke Br. eascu, UaIr. ace, arii ex *naghran, eas, azr, Ormant effoit, chongbail, uimir) and that they may accordingly represent an Old

pronoun or*

that

they have

each

lost

Celtic no-r,
'

just as the Ir. hor,

bar, far

'

you,' /or-n, hor-n,

far-n
pro-

your'

may stand,

respectively, for vo-r

and vor-n.

The most

bable hypothesis

is

that the copyist has misread the n of unet as r.

54. //3zWc?M (gl. transtris).

Prof.

grapher

may

have regarded

transtra as
It
is

Bugge thinks that the glossomeaning " quaecunque

has also been conjectured,
pi.

tigna in piano transversum posita."

with some probability, that libiriou

dibir,
it

a mistake for dibiriou

of

now

dibr.

Corn, diber

(gl.

may be

a mistake for Ubiniou

is

= the

sella),

W.
Ir.

dibyr.
'

Possibly, too,
galleys
',

liberna

which

occurs in LB. 119 b, and

a loan from the Latin libiirna.

55. deleioii (gl. antemnarum). Plural of dde, now dlez 'vergue ou antenne,' hors de Leon dl,' Legon., Corn, dele (gl. antempna) rod,' Goid-. 176, del, Corm. s. v. Caindelbra. Z-. 1070, Ir. detl
' '

56. guinodroiiou

(gl.

plagae).

A compound
venaior)

of guinod, bor-

rowed from

Lat. venatus (as IMid.

Br. guiznezl, giimhezl,

Mod.

Br. gwnaer, givinaer

from
ros,

Lat.

and

of roitou, IVIid. Br.

roedou Z-. 98, 287,

(gl.

now

rouejou, pi. of roit, roet Cath., Corn, ruid

rethe),

Mid. Corn,

D. 54,

W.

rhivyd,

all

borrowed from

Lat. rete.
57. ^zz;f/o
.i.

lanceae uenatrices

(gl.

uenabula),pl. oi giiinucl,

from the same root as guinod, No. 56. But the suffix -ucla is Ir. miiinl. Celtic and is foundalso in W. mynwgl collum," Z-. 820 58. utgiirthconeti (gl. obnixus .i. perdurans .i. contra nisus)
'

the context

is

Ille lovis

monitis

immota tenebat

Lumina, et obnixus curam sub corde premebat. This seems a mistake for utgurthconetic, the part. pret. pass, (here used with an active meaning) of a verb *ut-gtirthcotiam,
* The following examples 38 (Goid. 2d ed.64, where it
will
is

suffice:

con-or-tin6ltar
:

(gl.

locemur) L. H.
isin

wrongly explained)

ragmuidne at degaid

miiir con-or-bditer

and

'

we

will

go

after thee into the sea, so that

:

we may-

After the particles no and w: norbe drowned there,' H. 2. 16, col. 371 forraig do gleo garb gl ror-briss is ror-buaidre, LL. 50 b. I. Compounded winithe ^re^. do : no-dar-be-ne .i. biaid linne, 'we shall have,' O'Cl. Gl. The corresponding form of the 2d pers. pi. bar, bar is exemplified in Rev.
Celtique
b,
III,

95: do-bor-ficba, LU. 15

a,

a,

ro-bor-ficba,

84

a, ro-bar^ciered 84.

ar-nch-bar-accaister 85

do-for-fuc, ro-bar-bia, LL.

197, a.

2,

do-bar-

beraid,

LL. 46, b.

2,

ro-bar-tinoil

Lecain, col. 647, to which

,

may

LB. 8 a, do-bar -ruachtadar, Leb. Buide be added ro-far-cruthaigfe, LB. 184 a. b,

ro-bar-dimicnigsebar LB. 184 b.

12

Old-Breton Glosses.
prepositions

compounded with two

ut for ud,

Ir.

ud

(in uccu

ex ud-gu), od, Z^. 878, 885, Skr. ud, in Gr. va-rspos ex yS-xspo-. Goth. -2/, O. H. G. W2, N. H. G. aus, and gur/k Z\ 682,905, O. Wgurt, Z2. 1057
Prof.
'

Ir./?r/, frith Z^. 875, Lat.

it

j'^rj-MJ.

As
'

to <roa
elevare,'
(gl. lac-

Bugge puts
'

with the
It

W,

cynu

'

surgere,' erchynu

exaltare,' Z^. 92, 895.

vigilo,' as

seems

to occur in

anguoconam

to,* not
etic

printed in Z^. 1054).

in

As

to the termination in

(the regular

form

Welsh)

cf. dehlouetic infra

No. 67 and the

pi. craseticion (gl. spicis, leg. spissis) in

Lux.

II. GLOSSES IN
59.

AMALARIUS.

amsauath (gl. uicarius). The context is: "quasi ergo ant^ [Christi iudicem sic ante sacerdotem quia uicarius Xti. dni. Domini] est." The root of this word is sta (Curtius, No. 216), whence stam (Ir,
se-ssam) sam, sav
lev.'

now

sao

'

posture d'un corps qui est debout,

'

The

termination -ath for -a/ as in lagat

oculus,' Z^.

839

the

am- here signifying am-ryw, am-gen, 7?. 897.

prefix

variety or interchange, as in

W.

am-liw,

60. blinder (gl. segnitia)

= Mid. W. blinder
inerte)
lassus,'
'

'

fatigatio,' Z^.
(gl.

829:

a derivative from blin

(gl.
'

Lux.

pi.

blinion

inertes)

Lux.
'

W.

blin

'

fatigatus,

defessus,'

which Prof. Bugge

regards as a participle pass, in na (Beitr. VII 67) Skr. gldna erschopft." For the formation by der cf breinder putredo,' berrder
'
.

'

brevitas,' etc.

Zl

829. habihs).

61. camadas (gl.

The

context is:

"qui secundum

uerba sancti Gregorii semet ipsum metitur ipse habilis est." This is the O. W. cimadas (gl. par), Kuhn's Beitr. VII 390, now cyfaddas,
the
Ir.

comadas

'

fitting,'

'

meet,' Z.^ 994,

from com- and adas,

(gl.

W.

addas.
62.

Non minus

nahulei uel

int

coucant

Nihilominus in
is

caeteris operibus

quantum
in naouac'h

segregetur).
'

In nahulei^ the nahu

the negative

seems now naou

but
I

nanmoins,' where na

nid ZM-llai), Corn, le, Ir. laigiu is The adverb int coucant is lAao-crcuv. Mid. Br. cougant certainement,' Poemes Bretons du Moyen Age, W. j/ geugant, ceugant certo, certus,' addawyn 3, 248, 266, 279. geugant certo promittere,' ar peth yn angheugant re incert et
lei (like Ilai in

cannot explain the -hu.

The

W.

'

'

'

'

Mhik,' ceugant
int

= ent, No. 63, is = the Greek vr

In Welsh and Cornish
it

yw angaw

'

CQTi2i

est

mors,' Davies.
*

The

prefix

like

'

in avTihvXos, avr/osoj,

ccvtIttuis.

assumes the form in,yn, Z.^ 61 5.

* Lactare, deficere in pondere,

Ducange,

Old-Breton Glosses,
III. COLLATIO
63. enterafib (gl. ambit).
64. entcrafho heuanemdoguot (gl. inopportunius se ingerit).

13

CANONUM,

First Copy.

The

context

is

" sic

is

qui ultro ambit uel inopportunius se

ingerit procul dubio est repellendus."

W. chwy/ movti,' In the former gloss /<?>, the modern /wz;, is movere Davies), from a ox chwif {chwifiwr 'vagus,' chwimio groundform svimit, the infected b being written for infected m.
' ' '

The

Br.

from sv

is

to

Corn, di-fun
Z.2 905), /^r

sleepless.'

be compared with that oifariel, Jubuenn, The verb is here perhaps compounded
en-

with three prepositions (Z.^ 908),

(W.

en-,yn-, Corn, in, en, an,

(=

Corn,

ier,

der,

Z?

681), and a

(= ^.ac,Z?

673).
tes)

The somewhat similar Corn, entre in entredes {=z entre + en { tre. (gl. cauma i. e. xcwfjioc) is But Prof. Bugge reads

the gloss thus

int,
*

enter a fib
ter,

and explains
'

enter as
'

an adverb
impetueux,'

= ent =
W.
taer

No. 62, importunus/

now

tear
'

vif,'
',
'

'

prompt,'

'instans,'

urgens

from stagra
stechan.

= O. Bactr. tighra
gloss, ent crafho
la

which he conjectures may come keen and cognate with OHG.

'

The

a he thinks the relative

(cf.

a arecer

gl. cianti,

In the
in

latter

(leg.

crafhoch P)

-which occurs

Lux.)

Vicomte de
is

Villemarque's, not Mr.

'

Bradshaw's, copy

be an adverb translating
298,
the equivalent of
'firmus,'
'

inopportunius,' just as
fortius
'
:

the positive
'

must craf would be =

W. yn gallach Z?
is

the

^.craff

whence

craffu

fortiter

invadere'Z.^ 163,

Corn, craf

(gl.

avarus).

The h
So

in

craf-h-o
'

due

to the

accent on the following vowel.

Gurm-h-ailon, supra No.

8.

M.J, 21b, and de-h-urin, coil-h-aam, supra No. 34.

diutius,'

in pel-h-och

So

also in the loan-words ca-h-un caminus, com-h-azr

commater,
guiuf-h-er

e-h-oc

esocem, guifi-h-ezl venator, gou-h-iri

melin-h-ezr

vagi'na,

vivrra,

molinarius,

maing-h-ez

masnada,

pro-h-em

prooemium, rem-h-et remedium, salin-h-ezr salinarium, vail-h-ant valntem. For Welsh and Cornish instances of the evolution of h by the accent, see Rhys, Lectures, 2d ed., pp. 230, 231. In heuan em-d. we have, first, ev the Mid. Br. eff, Z? 372, now hen, O. W. em, Skr. ama, and then an-ein the Mid. Br. en-em in d-en-em dtffen ad se defendendum,' M.J. 179 a, where em is the common prefix used to form reflexive verbs, Z.2 899.
'

doguot
tive of

is

either for dogout, later dougo, the

3d

sg. pres.

conjunc'

dougaflo be compared with

like
*

and oe(t) digonwy

'servet,'

guledichuy

'

Welsh forms in -wy(t) dominetur,' cotvy i. e. coddwy laedat/

early
'

faciat,'

carwy 'amet,' rodwy

det,'

syllwy
in

'

videat,'

catwy

and

creddoe 'credat'

(Evander Evans

Arch. Cambr.

14

Old-Breton Glosses.

April 1873, p. i48),or the beginning of some verb compounded with the two prepositions do and guo, Hke doguomisuram, doguorejiniavi,
Z.2 907.
If the latter

conjecture be right,
:

we may perhaps supply

is
:

the wanting letters thus

dogiiotouc,

where touc for zouc

</i9m^<7/"
'

the
cf.

third sg. pres. indie, act. of the verb

dodocetic (gl. inlatam)

porto/

Z?

583

Lux. 64.

65. douretit atigncit (gl. turpi).

The

context

is

" diaconos

non multo uino

words appear
to

deditos

non turpi lucrum sectantes" and the Breton

'

mean

foeditas turpis.'

seems an abstract noun foedam), see No. 23. *engrt, Lat. aegretum from aegreo, As to angriiit it is perhaps as O. Fr. engrot is found from Lat. aegrotatiis, as the O. Fr. heingre (Br. hacre obscenus,' turpis/ Cath.) from Lat. aeger. The phrase douretit angruit would thus correspond in meaning with the O. W. dafraud atuis foedum opprobrium,' supra. No. 23.

As

to douretit

the ms.

has dourkit

it

formed from the

adj. douret

^^daureth

(gl.

'

'

'

66. id costadalt (gl.

*

aeditui aeclesiarum)

here id

stands for

and costad-alt is an improper compound (like mel-gahr supra, No. 2) o costad, a loan from Lat. custos, custodis, and alt Ir. alt .1. teach 'house,' O'Dav. 54, whence ailtire \. saor denma iighi, ibid. Prof. Bugge compares the Med. Latin custos ecclesiae aedituus, also custos basilicae, sacrarii, altaris (DuCange), O. Fr.
id
est,'

= =

coustre

'

sacristain,'

OHG.
et

custor,

NHG.

kster.

67. dehlouetic

(gl.

accommodata).

The

context is:

" uox

lectorum simplex est

clara pronuntiationis genus (uel generi)

accommodata." Vicomte de la Villemarque suggests that this W. delwedig, pret. part. pass, of delwi to form,' gloss may be This seems to be right, though it leaves the to fashion.' unexplained. Delwi is a denominative from delw figura,' forma,' (O. W. delu gl. numismatis Juv. 80), Ir. deld. In Cornish it is found only as an adverb dell deveras how it dropt,' P. 22 1, 3 (where how the second / is an assimilated v), dell-o lesus dystryppijs del-ma 'hoc modo,' sic,' Z.2 734, Jesus was stript,' P. 177,1 The context is " multi clericorum ieiu68. amuoet (gl. fastu).

'

'

'

'

'

'

'

nant fastu superbie ex propriis suis nihil largientes egenis." In amuoet, better amhuoed, we have, according to M. de la Villemarque, the O. Breton
reflex of the

Mod. W. ymchwydd
'

'a swelling

Pughe, who cites the sarcastic adage tri pheth mawr o ymchwydd ymhorth ar waed gele, horen a rhyfelwr three things a leech, a tike* great with self-swelling by feeding on blood
of one's
self,'
: :

* Sic.

Davies explains horen by

"

'

pediculus suillus

'

so possibly by

'

'

tike

Pughe meant

tick,

Fr. tique.

'

Old-Breton Glosses.
and a
used
warrior.'

15
arti-

Here,

if

M. de

la

V. be right,

(\W.ym-)

is

898) and {Ii)iwet is the ancestor of thelNIid. Breton huez, Cath., now chouez 'souffle,' whence s'enfler,' grossir,' W. souffler,' huezaff, Cath., now choueza chivydd tumor,' chwyddo tumere ;' or is the W. cognate chwythu
to express reflexive action (Z.' '

'

'

'

'

to

blow

'.?

But

it is

possible that

we should read a

nitioet,

regard-

ing a as the preposition used to express the ablative (No. 7) and W. f/myad pi. viwyadau auctiones, niuoet for mouet as augmenta.' * The context is: "propter 69. ahscenid. .i. caul (gl. piacula).

'

'

piacula regum."

Here abscenia
Ir.

is

for

obscena and caul

(gl.

is

= W.
Ml,

cwl
cited

'

culpa,'

'

peccatum,'

col,

gen. tn chuil

by Muratori, Antt.
guanorion

Ital. III. col.

871

au being written
0. is
:

piaculi)

for

u, as ao in ann-aor, supra,
"JO.

No,

5, is

written for

(gl. istriones).

The

context
'

" impudicos

et

non nutrire." PL of guanor 'histrio,' scurra,' a derivative horn guaan (gl. scurilis) infra, No. 81, which seems identical with W. gwann debilis,' Ir. fami, Lat. vanus from *vac-nus. Compare the W, dyn gor-wag scurra.'
istriones
*
'

71, torlehericii

(gl,

phitonistarum,
et

leg,
et

pythonistarum)

The

context
tionibus
'

is

"

magorum

phitonistarum

augoriarum superstiof tor, better torr

Ir,

non intendere."
leberieti, pi.
'

This

is

compound

venter

'

Cath., Corn, tor (gl. venter), of leberiet

W, tor,

tarr (cf, tar, supra,

No, 38) and

'

'W. llafariad
in

'an

uttering,'

a speaking.

For the termination

is

-iat see Z.

840,
the

meaning

of the gloss

thus

'

ventriloquisms.'

Compare
:

The Mod.
liceat

W,

bol-la/ariaeth.

The context is "Episcopo (gl. agipam). commendare uestimentum quo utitur et agipam et taxam."
72, gulcet
is

This
Cath.)

the

]\Iid.

Br, golchet {golchet da gouruez

ir cilchetou
(gl.

'

coete de

lit,'

O.
(gl.

W,

cilchet, pi.
gl.

vela) Z^, 1056, O. Corn, cilcet

Ir,

tapiseta,

stratorium) Z^. 1063,

culcita.

colcaid,

Z.2

802,

all

borrowed from Lat.

cf,

For the weakening of

f to

in anlaut

gant, gueffret and goural.

73, ar

uuo art hut

Ix.

aruiioart) stands for

with the O.
root
is

{g\. uos fascinauit). Here ar uuo art (leg. ar-guo-garth, a /-preterite, to be compared ad-ob-ra-gart (gl. uos fascinauit) Z? 455, The

Curtius, G.
infra

and other words collected by For the infection of the ^ cf. bu-orth No. loi, and the examples in Z.^ 202, 103. For the developyr^puM

GAR, whence
E. No.

133.

ment

of

meaning

cf. yr^s

(root

GU),

incantare,

Curtius, G. E.,

* W. WTwyrf 'that is swelled or puffed up,' Pughe, is (as Bugge points out) still nearer our muoet but with this meaning it is at least doubtful.
.

Old-Breton Glosses.
(gl. incantator).

No. 642, and Corn, vur-cheniat

ar-fo-imim
'

The

Irish verb

recipio'

is

similarly

compounded

with the prepositions

ar and/b

= guo =

vtto.

hui

is

the pers. pronoun of the 2nd pi.

Mid.

Br.

c'houi, Z2. 371,

Ir. -si, sib

Corn, why, O.

W.

hui,

Mid.

W.

chwi.

huy, now From *m,

*si-svi-.

IV. COLLATIO
74. aiul (gl. ultro).

CANONUM,
is

Second Copy.
a,

This adverb
subst. iul,

formed from the prep,

'

Mid. Br, youll, M. 10*, eoll: am. eoll a ma volont,' Cath., now ioul. Corn, awell, awel desire/ W, Root AV, whence Lat, avidus, avarus, avere, ewyll, ex *avilla.
supra, No. 7
'

and the

and possibly
(gl. gratis).

(as Siegfried thought) the Irish deolid 'gratia/ indeolid

75. aruanta

the context

is

"Sic

is

qui ultro ambit uelinopportu-

Mr. Bradshaw says, nius se ingerit procul dubio repellendus." " The word aruanta (I am not quite satisfied about the second

margin opposite the line beginning inopThere is no referring mark." ? It is I conjecture that aruanta is a gloss on "ultro ambit." compounded with the prep, ar, 7?. 900, and the -uanta may possibly be for huanta (cf amuoet if for amhuoet) the 3rd sg. pres. indie,
letter) is written in the

portunius.

Does

it

refer to ingerit

act. of

a verb cognate with

W. chwannawc
Ir.
i,

W. chwennychu

'desiderare/ chwant ==

desiderans/ Z^. 153, sant 'desiderium' and O.

'

Br. couhuantolion gl. andrivenereis

right reading of the

e.

cupidi

(if

this

be the
508, to

Luxemburg gloss). For the third Mid. W. kanhatta, teruyna and other forms cited

sg. in -a, see

Z^.

which may be added penitra (gl. tractat) and the following Evander Evans * doluria late Professor collected by the 'dolebit/ eheta 'convolabit/ cerda 'procedet;' these from the
:

oldest copy of the

<

laws

'

guada
:

'

denies/
'

palla

'

fails/

gnaa
treidia

does

'

these

from Cynddelw
'hastens/

puylla

considers/
?ix\di

'penetrates/
'

bryssya

atveilya

'decays;'

yd

aa

goes

'

from Llywarch Hen.

clericos

The context is: " non oportet quibuscumque spectaculis incaenis aut nuptiis interesse, " whence it would seem (according to Mr. Bradshaw) that the Latin word intended was encaeniis/ which must mean secular festivals,' Here inuanetou (if this be the
76. inuanetou (gl. incaenis).

sacerdotes uel

'

'

right reading)

is

the pi. of inuanet.

The

in

seems the prep,

in-

used as a

prefix,

Z?

905, but infecting as in

II, in

Mod. W. yn/er-=

* Studies in

Cymric Philology, No.

Archaeologia Cambrensis, April

1873, p. 147-

Old-Breton Glosses.
Ir.
'

17

to

inbher influxus.' The -uanct (from hanet ? nianet ?) is obscure me. Is Ir. banessa (gl. nuptiae) cognate ? or is the uninfected form Vianet cognate with Lat. meti-sa, as banquet with banc (' c'est

ainsi qu'en

AUemand

tafel possde la fois le

?

sens de table et

celui

de
of

festin,'

Brachet)

Or,

lastly,

should we read in uaretou

(gl. in cenis)

pound
et
*

and regard uaretou as the pi. of uar-et, a comuar 'evening' (Corn, uher, uer, uar, W. ucher) and a meal Mid. Br. eth corn,' W.jyd, Ir. It/i Skr. pitu ? It
'

'

is to

be hoped that a new collation of the ms.

is

will justify this

sug-

gestion, which

due

to Prof.

Bugge.

77. dadluo (g\. SLXitTopas).

et
is

The

context

is

" Clericus per plateas

antropas

nisi certa necessitate

'

non ambulet."
'

Here 'antropas'
Festo
et aliquot
viri, 01

miswritten for

andronas,' ace. pi. of ajidron.

scriptoribus Latinis Andron, compitum, locus publicus ubi

avtpsc, invicem confabulantur,'

Ducange. Dadluo (=: dadlou,

infra,

No. 79)
Eutych.

is
s'',

the pi. of
8;
'

dadI (ms.

dadlt) (gl. curia), dadl (gl. concio)

Ir.

O.W.
for
-

datl (gl. foro), datlocou (gl. fora) Z.2 1055,

(gl.

ddl 'curia,'

forum,' Corn, datheluur

is

concionator).
(gl.

The plural
scropibus).

ending

uo

ou, as in

O.
5"^.

W. crtimmanhuo
is

An

older Breton form datol (for datl)

(gl. lego)

preserved in the verb

is:

datolaham
78. clou

Eutych.

(gl.

acitamenta).

The
and

context

"unus

uendidit

acitamenta eius in oblationem ecclesiae dei."

the space between 'acitamenta'
Prof.
cloeu
'eius.'

The
I

gloss stands over

it.

cannot explain

Bugge
'

conjectures that clouxsioxcloow^X.oiclo

= clavusi^.

clavi,'

Z?

'nail,' Ital. agiito,

285) and that acitamenta is for acutamenta (acutus Placidus XXI 91 ed. Deuerling.)

79. endlim (gl. fenus).

Mid.

W.
'

ennill,

Y Seint

Greal, 42, 196,

emolumentum,' Ir. indile .i. torviach 'increase,' H. 3. 18, p. 71, col. i, and see Cormac, In Irish indile or innile also meant cattle techit ass p. 96. For the termination fochtoir ind innile 7 in bdachaill, LU. 26a. of endlim, cf. diprim, erchini, molijn, 7} 821, stlinnim, Kuhn's Beitr. IV. 392 and coriin No. 44.
jyfmill
'

now

lucrum,'

'

qusestus,'

'

:'

v. COLLATIO CANONUM,
80. dadlou (gl. andronas). 81.

Third Copy.

guaan

(gl. scurilis).

See above. No. 77. See above, No. 70.

'

82. nith

(gl.

nepta).

thia filiaque uel sorore nepta

fratre uel sorore,' is
nith,

context is: " Clerici cum matre uel tantum uiuant." Here nith neptis ex Mid. Br. nyz, Cath., Corn, noit (gl. neptis), W.

The

O.

Ir. 7iecht (gl.

neptis) Z." 68.

83.

mod

{g\.

gomor).

The

context

is:

"Sciendum quantum
B

i8
est

Old-Breton Glosses.

pondus primitiarum .i. gomor." Here gomor is the Hebrew homer or chomer, a measure for things dry, and mod, like Mod. W. The Ir. muide churn,' 7n6dd, is borrowed from the Latin modus. W. buddai, and Mod. Br. mz seem from modius, Fr. rnuid. So in Lux. rannou (gl. patri84. rannou (gl. cHmatibus). Hence pars,' Cath., W. rhann, Ir. rann. PI. of rann monia).
'
'

rannam
86.

(gl. partior)

Eutych.

4.

85. cunrunt (gl. uorticem montis).

a</zi?/ (gl.

passae).

The

context is:

"cumque dauid

transiet

paululum uorticem
\sic\

montis apparuit ei siba ... cum duobus asinis qui honorati erant cctis panibus et .c. alligaturis uuae passae."
'

summitas,'

Here cunrunt seems a compound of cun (W. cwn altitudo,' Z? 92) and runt for rund now rond rotundus,' nt
'

being for nd, as in

admet
is

= W.

cantoell, confuntaff, respontas, etc.

add/ed
Ir.

'

maturus,' Mid. Br. azff, Rev.

either ex mati

Celt.

i.

399, Corn, arvez,

ad, T?. 897,

abaidh (ex ad-vati, ad-mati), from the prefix

is

and met, which

and connected with

Lat. mat-urus, Mat-uta, matu-tinus, or

(as Prof.

Bugge

thinks)

cognate with Old Welsh

87. controliaht (gl.

metetic

'

reaped,'
is

et-met

'

retonde,'

Mod.
iur-

Br. mdi, midi, just as the English ripe

controuersiam).

cognate with reap. The context is " De

:

gatoribus

quod per controuersiam cuncta

I,

faciunt incerta."
:

Here

controliaht (for controliacht, -iact, or for controliath

cf

W.

dranoeht.

Laws

2,

27)

is

a derivative from *control, Mid. Br. contrel,

borrowed from Lat. contrarius, whence also Mid. W.

'the Devil,'
in

kythreul

gwynt

Seint Greal 75, kythreul, Z*. 819, cythrawl a contrary wind.'

'

now

cythrawl, as

88. latic

(gl.

agipam).

The
[

context is:
]

"

commendare
Here
loit

uesti-

mentum quo
'

utitur et

agipam

taxam."

latic is

the

equivalent of gulcet, No. 72 supra.

It

must be a loan from Iddix

in dia
s. v.

a counterpane.'
'

white blankets
of latic, says

Hence also the Ir. hit Cormac (Mac Firbis' copy)

find

'

two

cermnas.
little

The a

just destroys the letter.

Mr. Bradshaw, may be 0, but there is a For Br. a == Latin 5, cf.

hole which

89. gutric (gl. defer). uel exigendus defer."

of differ.

The

context

is
:
'.

No. 66. " Si debitor inrogandus

costad.
'

delays,

'

The margin says ffer, probably the remains Compare O. W. nouinn-guotricu segeticion nine blessed Mid. W. godrigyaw to tarry/ Y Seint Greal, 227, Mod,
'

W.
'

godrig

'

mora,' from guo and trie {trigo

'

'

morari

')

cognate with,

or borrowed from, Lat. tricari

hindrances,'
'

to

tricks,'
'

Prov.

trie.

make difficulties,' to trifle,' tricae Hence also the Mid. Br. trig in

the phrase hep trig

sans tromperie.'

Old- Breton Glosses.

90. aaltin
(gl.

19
''nee
ferula curare

ferula).

The

context

is:

meditetur quisque quod gladio percutiendujn." used to indicate the ablative (as in Nos. 7, 17,
Br. autenn

now

aolen

'

rasoir,'

'

couteau

'

is

Here is the prep, 4O and aliin, Mid. the O. Corn, elinn (gl.
rasorium acutum) Ml.
is

nouacula)

Z?
amal

1062,

W.

ellyn,

O.

Ir.

allain .1. scian bearrlha,

O'Dav.

54,

in n-altain n-ilh (gl. sicut

col. 301.

The glossographer

here, as elsewhere,

not exact in

his renderings.

91. 92.

aceruission (gl. hirsutis).

cunnaret

boestol (gl.

beluina
:

rabies).
alios

The

context

is

"alios hirsutis serra dentibus attriuit

ungula sparsit
minuit.''

alios

armato ferro insulcans beluina rabies morsibus detruncando comthe preposition used to indicate the ablative,

In aceruission a (No.
7,

is

supra) and f^'/'wmww stands for geruissioyi, pi. of gauiss, the

lost c of a,

being provected owing to the influence of the

'

which

is

still

found in the Luxemburg gloss ac-i(r)-riminiou. So in the Mod. Br. dk had ten hares (gad) and the Cornish drok-coleth from drok-\- goleth. With the gcriiiss thus obtained cf. W. gerwin
'

'

asper,'

Cath.,
cf.

asper, M. Br. garu, rigidus, a derivative from garw now garo, Corn, garow, Ir. garb. For the termination in -iss certus,' securus/ (Ir. the W. adjectives dilis, dylis now dylys
'
'
' '

'

'

diles),

hyspys
is

manifestus,
VV.

'

'

certus,' Z.- 834.

'

cunnaret
infected d,
'

=
'

cyndaredd

rabies,' the final

being written for

and the d in inlaut assimilated as in the Mid. Br. connar rage, Cath., whence conniryec rabidus = Corn, con(n)erioc (gl. rabidus). Mid. W. kandeiryawc, Y Seint Greal, 301, 418,
' '

now

cyndeiriog*

Celtic ciin-dara-s

These words must all come from some Old meaning hound-madness,' hydrophobia.' For
'

'

the nasal infection of

boestol is

see Z.^ 118, 205, 207, 901.

boest
bist.

an

adj.

formed from
(fit), Ir.

borrowed from Lat.

bslia,

whence

also

W. bwyst

93. dadlou (gl. curiae). 94. trot (gl. strutionem).

95. couann (gl.

See above, No. 77.

noctuam).

96.

trad'

The
*I
nait
'

think not trad,' says Mr. Bradshaw, (gl. larum). context is " Haec sunt que de auibus comedere non debetis
I
:

may
(gl.

take this opportunity of pointing out that the Cornish di-scoruurabies) Z? 1072, is to be explained by reference to the Br. curun

thunder'

Kspuvvog

for arxspuvvoc, Curtius G.E. 694, just

astheComisA

folter-guske
'

(gl.

freneticus)

fulgur.

'

Compare

is to be explained by the Br. /oultr, Fr. fouldre, the Greek Iju,/3^o'vt>jtoj, xipuwo^Xrji, and the Latin

aUonitus.

20

Old-Breton Glosses.
noctuam
et

[see Leviticus xi.15] strutionem et

larum."

The

first

of these birdnames

hke Corn, troet, Ir. struth, seems borrowed (W. ostruih is, like Fr. autruche from struthio or Prov. estrut. from avis-struthio and the Mid. Br. loirucc, Cath. is borrowed from
trot,

Fr. Vostruce,

couann
cauanntts,

is

now Vautruche.') now kaouen or kaouan hibou/ 'orfraie cauanus, Ducange, and many other words
'

:'

cf.

cited

Med. Lat. by Diez,

Etym. Worterbuch,
trad
I

s. v.

choe.

cannot explain, except as borrowed from trochilus, a kind

a
(gl. attacus).

of wren.

97. deuo

: : :

This

cannot explain.
of the

98. biunrun (gl. epimachus), the

of grasshopper.

<{;joj,a;)(^of

LXX,

a kind

attacus atque epimachus For biunrun we should certainly read (according to Prof. Bugge's conjecture) bianran, a comW. bychan* little and ran borrowed pound of bian for bihan

The

context
.

is

" comedere debetis

xi.

aclocusta".

[Leviticus

22].

'

'

from rana, which in German dialects (Grimm, Deutsches Worterbuch, s. V. Frosch) is called grashiipfer.

VL COLLATIO CANONUM,
99. saltrocion
tentes
(gl.

Fourth Copy.
is
.

graciles).
.

The
.

context
garrulae

:" Sunt
uagae
.

aliae peni-

quae

sic uiuere

uolunt uitiosae

fabulosae

graciles nihil

commodi

praebentes
'

aliis."

The word
cion is
'

glossed must be

uitiosae,'

not

'

graciles,' for saltro-

the pi. of an adj. saltroc, derived from saltr,

' '

nom

saotr
salo

salet,

corruption,

'

a loan (like Fr. sale) from the

OHG.

triibe.
Ir.

halou (gl. stercora) Z,^ 1063, and salach, seem cognate with salo.

The O. Corn,

W. halawg,
:

The context is 100. guohethe (gl. probum, var. lee. pravum). " Patricius. Non oportet indices tam ueloces esse in iudicio donee
sciant

quod prcbum at

quod dictum

est.
'

Noli iudex esse cito."

For 'probum' we should perhaps read

*

probrum,' guohethe seems

The

Irish

cognate

is

becc

(W.

bach),

which Prof. Bugge, with much

'

probability, regards as a very old loan from the vulgar Latin piccus spitz,' whence also Ital. piccolo, Sp. pequeno 'little.' Other instances of b for/) in
Ir. biail, W. bwyell hatchet from pialla Diez II 914, Ir. Rhys puts biixora pix, h. bac irom pacetn, Br. baradoes from paradisus. bach with o"jU,J)tpo'j {Rev. Celt. II. 189) and seeks a trace of an earlier becc,

Celtic loanwords are

'

'

anlaut zb (from sm) in the fact that in North Wales bach forms a remarkThus geneth able exception to the initial mutation of feminine adjectives. bach a little girl,' afon bach a small river,' not /ach, as might be ex' ' '

pected according to the general rule. Bugge justly observes that this peculiarity is explained by his assumption that bach originally began with p.

Old-Breton Glosses.
the

21
superlative of drouk

mod.

givam 'pessimus,'
'

the irregular

Corn, gue/ha, D. 1130, gueze, P, 196, 2, W. gwaethaf: cf. Mid. W. gwaethau to make worse ', Y Seint Greal, 141. The insertion of A between the elements of the diphthong oe for oa (cf goazhat em'
.

pirer,'

Cath.)

is

curious

= W, bahet = W. cain- and

hivyell,
(rit)

we

find

it

also in the Corn, ^a^,?//' securis,'

'
'

(gl.

in the

p.

loss of the final

232) and tranoheth m is remarkable

W. baedd boar, and chahen W. Juihahelo (Rhys, Lectures, 2d ed,, 'trans noctem,' 'mane/ Laws 2,1,27. The
aper^
;

a similar loss

is

noticeable in the

Old Cornish ms. Bodl. 572

Zeuss^
(gl.

1060-63 ^s Ol"^

which are printed in Welsh), where we find dowomisura mi

(the glosses in

compensabo).*

The context is: " canis peccorum loi. buorth (gl. bouello). quodcumque mali fecerit in bouello uel in pascuis " (his owner Bovellum idem quod bovile, in to make good the damage). is Canonibus Hibern. lib. 51, cap. 5, DuCange. Our buorlh would therefore mean a cowyard,' and is compounded of bu, Ir. bo haie, clos,' Corn, garth, garth bos, /Sou? and gorth now garz garden, M. Br. liorz, W. lluarth pi. in luworth, lowarth The Welsh equivaluird (gl. horti), Ir. gort xP'^i' 'tortus.' lent of our buorth is buarth, which Pughe explains " a cow-yard or
'

'

'

'

'

'

inclosure where cows are turned to be milked

cattle
;

a place to fold

a fold," and which he illustrates by the adage gwell buarth

hysb nag un

gwag

" a dry dairy

is

better than

With
arid
to

this

proverb the present essay

it

may

most readers,
full of

is,

thanks to

an empty one." though Professor Hagen and Mr.

fitly

close, for

Bradshaw,

new material

for Celtic philology.

* For this gloss, hitherto unprinted, I am indebted to Mr. Bradshaw. have had to represent by in the Anglo-Saxon form of that letter. Other Old Cornish glosses, hitherto unprinted, in the same codex, are cennen (gl. membra[na]) and genimoii (gl. saphero et exsmaragdo.) A new Old-Breton Mr. Bradshaw gloss from the Luxemburg codex is luson (gl. tramitem).
1

equates this with the

W.

Hyson

pi. of

llws

'

track.'

INDEX
a rel. pron. No. a prep, cum abl. admet 86.
ail 9.

63.
7, 17,

controliaht 87.
42, 90,91.

corcid 25.

corim 44.
cornigl 26.

a-iul 74.
alt 66.

cosmid
cos 35.

35.

costadalt 66.

akin 90.

amsauath 59. amuoet (?) 68.

an- 64. angruit 65.

couann
couarc

95.
12.

coucant 62.
craf, crafho(ch) 64.

annaor
aor
5.

5.

cronn

17.
17.

cronnmain
ii.

ar 73, 75-

Arbedoc

aruanta 75. aruuoart 73.

attanoc 34.

cunnaret 92. cunrunt 85. dadl 77, 80, 93. daureth 23.
dehlouetic 67.
dele 55.

bianran

98.

bidite 33.

deuo

97.

blinder 60.

disc 30.

bodin 43.
boestol 92.
bro, brool 45.

do

49. douretit 65.

du

II, 41.

buc

13, pi.

bocion 27.

duglas 41.
duliu
1

buorth loi.

1,

camadas
caubal
caul 69.

61.

elestr 3.

18.

emdoguot
en
6, 15.

64.

64.
79.

cauuel, cauell

endlim
ent 64.

cepr 42.
cest 20.

clehurin 34,
clo 78,

en-ter-a-fib 63. ercentbidite 33.

esceilenn 52.
50.
et (?) 76.

cnoch

coarchol 46.
cocit (cecit ?) 37.
coll, coilin 16, 32,

ethin

8, 31.

fib 63.

gabr

2.

conam

58.

geruiss 91.

24

Index.

glas4i.