Sacred Texts  Myths/Sagas  Celtic  Index  Previous  Next 

p. vii


THE old-Irish tale which is here edited and fully translated 1 for the first time, has come down to us in seven MSS. of different age and varying value. It is unfortunate that the oldest copy (U), that contained on p. 121a of the Leabhar na hUidhre, a MS. written about 1100 A.D., is a mere fragment, containing but the very end of the story from lil in chertle dia dernaind (§ 62 of my edition) to the conclusion. The other six MSS. all belong to a much later age, the fourteenth, fifteenth, and sixteenth centuries respectively. Here follow a list and description of these MSS.:--

By R I denote a copy contained in the well-known Bodleian vellum quarto, marked Rawlinson B. 512, fo. 119a, 1-120b, 2. For a detailed description of this codex, see the Rolls edition of the Tripartite Life, vol. i. pp. xiv.-xlv. As the folios containing the copy of our text belong to that portion of the MS. which begins with the Baile in Scáil (fo. 101a), it is very probable that, like this tale, they were copied from the lost book of Dubdálethe, bishop of

p. viii

[paragraph continues] Armagh from 1049 to 1064. See Rev. Celt. xi. p. 437. The copy was made by a careful and accurate scribe of the fifteenth or possibly the fourteenth century. The spelling is but slightly modernised, the old-Irish forms are well preserved, and on the whole it must be said that, of all MSS., R supplies us with the best text. Still, it is by no means perfect, and is not seldom corrected by MSS. of far inferior value. Thus, in § 4 it has the faulty cethror for cetheoir; in § 25 dib for the dissyllabic diib; in § 61, the senseless namna instead of nammá. The scribe has also carelessly omitted two stanzas (46 and 62).

The MS. which comes next in importance I designate B. It is contained on pp. 57-61 of the vellum quarto classed Betham 145, belonging to the Royal Irish Academy. I am indebted to Mr. P. M. MacSweeney for a most accurate transcript of this MS. When I had an opportunity of comparing his copy with the original, I found hardly any discrepancies between the two. B was written in the fifteenth century, I think, by a scribe named Tornae, who, though he tells us in a marginal note 1. that he had not for a long time had any practice in writing, did his task remarkably well. He modernises a good deal in spelling, but generally leaves the old-Irish forms intact. Thus we owe to him the preservation of such original forms as the genitives fino (13), datho (8. 13), glano (3. 12), of étsecht (13), etc.

p. ix

H denotes a copy contained in the British Museum MS. Harleian 5280, fo. 43a--44b. For a description of this important MS., which was written in the sixteenth century, see Hibernica Minora (Anecdota Oxoniensia, Mediæval and Modern series--Part VIII.), pp. v and vi. In this copy the spelling and forms are considerably, but by no means consistently, modernised. In a few cases H has preserved the original reading as against the corruptions of all or most of the other MSS. Thus it has cetheoir (4), muir glan (35), moitgretha (8), etc.

E is a copy contained on fo. 11b, 2--13a, 2 of the British Museum MS. Egerton 88, a small vellum folio, written in the sixteenth century. The text is largely modernised and swarms with mistakes and corruptions. By sheer good luck the scribe sometimes leaves the old forms intact, as when he writes órdi 14, adig 21, Ildadig 22, mrecht 24.

S is contained in the Stockholm Irish MS., p.p. 2-8. I am indebted to Mr. Whitley Stokes for a loan of his transcript of the whole MS. S is deficient at the end, breaking off with the words amhal bid atalam nobeth tresna hilcetaib bliadan (65). It is of very inferior value, being modernised almost throughout in spelling and forms, and full of corrupt readings, which I have not always thought it worth while to reproduce in ray footnotes.

L is the copy contained in the well-known MS. belonging to Trinity College, Dublin, marked H. 2. 16, and commonly called the Yellow Book of Lecan, col. 395-399 This MS. dates from the fourteenth century. It is of most unequal

p. x

value. The scribe, in his endeavour to make the original, mostly unintelligible to him, yield some sense, constantly alters in the most reckless and arbitrary manner. At other times he puts down whole lines of mere gibberish. A good instance of his method is the following rendering of the 34th quatrain:

Is ar muir nglan dochíu innoe
inata Bran bres agnæ
is mag mell dimuig a scoth
damsa i carput da roth.

As in the case of S, I have not thought it necessary to give all the variants of L. Yet in a few instances even L has by a mere chance preserved original readings abandoned by the other scribes, e.g. isa tír (6a), ind nathir (45), bledhin (62).

The six MSS. here enumerated, though frequently varying in details, offer on the whole an identical text, and have clearly sprung from one and the same source. For even the vagaries of L turn out on closer inspection to be mere variants of the same original text. Under these circumstances it was a comparatively easy task to reconstruct a critical text. In nearly every case the original reading was preserved by one MS. or another. Thus almost every form in my edition is supported by MS. authority. In the very few cases where I have thought it right to deviate from all the MSS., this has been pointed out in the notes. Still I am far from flattering myself that I have succeeded in restoring

p. xi

the text to its original purity. In some cases, fortunately not many, the readings of all the MSS. seemed hopelessly corrupt. See e.g. my remarks on dorearuasat, 48; aill erfind, 22; cach ági, at sáibsi ceni, 45. In other cases it is doubtful whether I have preferred the right reading. Thus, in to, I may have been too rash in adopting the reading of L, cen indgás instead of fri indgás of the rest. Considering the tendency of L to alter a less common expression into a familiar one, as well as the consensus of all the other MSS., I would now retain fri and translate it by 'with.' For this use of the preposition, cf. fri imḟochid, p. 85, 3. Again, I cannot claim that the text, as it now stands, represents the actual language of any particular period, containing as it does middle-Irish forms by the side of old-Irish ones. Such a mixture of linguistic forms is, however, not of my own making, but is an inherent peculiarity of most of our older texts, fully explained by the way in which they have been handed down.

But before I speak of this, I will try to determine as nearly as possible the time at which the Voyage of Bran was originally written down.

If we had any investigations into the history of the Irish language besides the excellent history of the Deponent lately published by Professor Strachan, it would probably be possible to determine with accuracy the time in which a particular text was composed. At present we must be content with much less certain and definite statements, often leaving a. margin of a century on either side.

p. xii

[paragraph continues] In the case of old-Irish, it is mainly by comparing the language of a given text with that of the continental glosses that we arrive at anything like a trustworthy conclusion, and this I propose to do in the present case.

There are a large number of forms in the Voyage of Bran as old as any to be found in the Würzburg glosses. The oldest part of these glosses, Professor Thurneysen, the most careful and cool-headed of observers, does not hesitate to ascribe to the seventh century. 1

I now subjoin a list of these oldest forms, leaving aside anything of a doubtful or unexplained nature.

First. as to sounds and their representation, the following archaic forms and spellings are noticeable:

Final e, early broadened to æ, ae, later a: sube, 8; comamre, so: móramre, 29: labre, 29: blédne (later blíadna), 55, 58.

Final i, early broadened to ai: adamri, cadli, 11; órdi, 14; crédumi, 14; also blédin (later bliadain), 62; adig (later adaig), 24; athir, 45, 57; i for infected a: Ildadig, 24.

Initial m before r: mrath, 9; mrecht, 23, 24; mruig, 9, 23, 24, 54.

ld for later ll: meld, 14, 39; inmeldag, 41.

éu for éo: céul, 9, 18, etc.

ói for later óe: cróib, 3; óin, 13 tróithad, 30.

Also, perhaps, b for f in graibnid, 23; airbitiud, 18; and oa for úa: sloag, 17 (R), cloais, 9, etc.

p. xiii

In the declension, notice the neuter nouns a rígthech, 1; a céol, 2; am-mag, 5; am-muir, 12; muir glan, without nasal infection later added by analogy with neuter o-stems, 17, 28, 30; fris’ tóibgel tonnat, 2; cusa cluchemag, 20; isa tír, 62, etc. The following genitives sing. of i-stems occur: glano, 3, 12; mora, 37; of u-stems: betho, 27; fedo, 42; fino, 13: datho, 8, 13; the datives sing. of o-stems: láur, 1; Braun, 2; the accusatives plural: rúna, 52; nime, 28: muire. 48; tedman, 21; the genitive plural: dúle, 44.

In the article the full form inna is of constant occurrence. In the poetry it is twice shortened to ’na in the gen. plur. (26, 30).

Among prepositions, notice such a form as dóu, 29, 32, 51; the use of íar with the dative. 26, 32; the careful distinction between di and do.

But it is in the verbal system that the archaic character of the language appears to greatest advantage. The distinction between conjunct and absolute as well as between dependent and independent forms is preserved throughout.

Present indicative, sg. 1: atchíu, 15--sg. 2: immerái, 37; forsn-aicci, 38; nad aicci, 19; nofethi, 49--sg. 3: mescid, 16: canid, 18; graibnid, 23; forsnig, 6, 12; dosnig, 12, 22; comérig, 17; tormaig, 18: foafeid, 22; immaréid, 33; frisbein, 16; frisseill, 59; forosna, 16; consna, 5; immustimerchel, 19; taitni (dep.), 6; tibri (dep.), 35; donaidbri; 17--pl. 3: lingit, 38; bruindit, 36; taircet (dep.), 14, 40; ní frescet, 18, 23 immataitnet, 4; taitnet (dep.), 40; taitnet 

p. xiv

[paragraph continues] (independent!), 8, 36; congairet, 7; forclechtat, 5; foslongat, 4; frisferat, 21; forsngairet, 7.

Present subjunctive, sg. 3: tróithad, 30; imraad, 60 ; étsed, 29.

T-preterite, sg. 3: dorúasat, 27 ronort, 46.

Reduplicated preterite, sg. 3: ruchúala, 20.

S-future, sg. 3: silis, 55; conlee, 51; adfí, 52. Secondary s-fut., sg. 2: rista, 30.

Reduplicated future, sg. 1: fochicher, 56; arungén, 57--sg. 3: gébid, 26; adndidma, 51; timgéra, 59.

E-future, sg. 2: ricfe, 60--sg. 3: glanfad, 28; dercfid, 55; ticfa (independent!). 26, 48; rothicfa, 49; móithfe, 52; fuglóisfe, 48; ícfes, 28.

Imperative, sg. 2: tuit, 30; tinscan, 30.

Verbal nouns: étsecht, 13, 24; óol, 13; imram, 17; airbitiud, 18.

The following passive forms occur: pres. ind. pl., agtar, 54; sec. pers. sg., atchetha, 12, 39; red. fut. sg., gébthir, 57; gérthair, 51; pret. sg., adfét, 29; atfess, 29; s-fut. sg., festar, 26.

As to old syntactic usage, notice the adjective and substantive attributes placed before the noun, 4, 13, 19, 29, 43.

Lastly, I would draw attention to the use of the following words as dissyllabic, though as most of them continue to be so used as late as the tenth century, such use is not in itself proof of great antiquity.

bíi, 9; bíaid, 50, 53, 55; bías, 27. Cf. Salt. na Rann, ll. 8021, 8202; Trip. Life, pp. 70, 22; 222, 4, 6, etc. But

p. xv

their use as monosyllables is far more frequent in Salt. na Rann. See ll. 835, 1076, 1599, 1951, 1952, 2043, 2047, 3275, 3320, 3353, 5046, 6255, 6325.

cía, 'mist,' 11.

criad, gen. of cré, 'clay,' 50, as in the dat. criaid, Salt. 7683, 7769. Monosyllabic in Salt. 394 (leg. criaid), 8230.

día, 'God,' 48. Cf. l. 18 in Sanctán's hymn:

friscéra Día dúlech.

and Salt. 1905, 2013, 2685, 5359, 7157, 7969, 8074. Monosyllabic in Salt. 649, 1917, 1950, 2742, 3121, 3308, 7976.

diib, 'of them,' 25; as in Salt. 375 (sic leg.), 437. But monosyllabic in Salt. 4975, 4985, 5461, 5417, 5869, 7704.

fia, 11.

fóe, 'under her,' 6.

óol, 'drinking,' 13. Cf. oc óul in the Milan glosses (Ascoli); d’óol, Salt. 1944.

úain, 'lambs,' 38.

It will be observed that the above forms are taken almost exclusively from the poetry. The prose, though it preserves a large number of undoubtedly old-Irish forms, also contains a good deal of what is clearly of middle-Irish origin, more particularly in the verbal forms. The use of preterites without the particle ro has been recognised by Thurneysen, 1 whom I mainly follow here, as a decidedly later phenomenon. It occurs in birt, 31; asbert, 62, 63 (bis), 64, instead of old-Ir. asrubart, and in a large number of

p. xvi

s-preterites such as fóidis, 61; gabais, 63; scríbais, 66; celebrais, 66; sloindsi, 62. We find dobert 2, instead of old-Ir. dorat, and dobreth 62, instead of doratad. The late cachain occurs three times (2. 32, 65), for old-Ir. cechuin.

Such Middle-Irish forms, which all MSS. without exception contain, show that the original from which our MSS. are in the first instance derived, cannot have been written much earlier than the tenth century. Bearing this in mind, together with the occurrence of the seventh century old-Irish forms side by side with these later ones, as well as with the fact that the poetry contains none of the latter, we arrive at the following conclusions as to the history of our text.

The Voyage of Bran was originally written down in the seventh century. 1 From this original, sometime in the tenth century, a copy was made, in which the language of the poetry, protected by the laws of metre and assonance, was left almost intact, while the prose was subjected to a process of partial modernisation, which most affected the verbal forms. From this tenth century copy all our MSS. are derived.

In conclusion, I would draw attention to the loan-words occurring in our tale. These are all of Latin origin. 2 They naturally fall into two groups, an older one of words

p. xvii

borrowed at the period of the first contact of the Irish with Roman civilisation, before the introduction of Christianity; a later one of words that came into Irish with Christianity. To the first group belong aball, 'abella'? 23; arggat, 'argentum,' 23, 14, 22; drauc. 'draco.' 13; dracon, 'dracontium.' 12. 58; fín, 'vinum,' 13, 14; fine, 'ab eo quod est vinea.' Corm., 43; port, 'portus,' 62.

Of words of the second group we find: cór, 'chorus,' 18; corp, 'corpus,' 46, 50; líth, 46, through Welsh llith from Lat. lectio; mías, 'mensa,' with the meaning 'dish,' 62; peccad, 'peccatum,' 41; praind, 'prandium,' 62; oceon, 'oceanus,' 25; scríbaim, 'scribo,' 66.


It remains for me to express my gratitude to those who have taken a friendly interest in the production of this little book, and who have in various ways given me advice and assistance; above all to Mr. Whitley Stokes, to whom I am indebted for many weighty suggestions, as well as for the loan of valuable transcripts; to the Rev. Richard Henebry, Mr. Alfred Nutt, and Mr. P. M. MacSweeney, and to my kind friends and colleagues, Mr. John Sampson, and Prof. John Strachan.





vii:1 An abstract and partial translation of the Voyage of Bran was given by Professor Zimmer in the Zeitschrift für deutsches Alterthum, vol. xxxiii. pp. 257-261.

viii:1 This note is found at the bottom of p. 57 and runs thus: Messe Tornae 7 ni fetur ca fad o doscriuhus oenlini roime sin, i.e. I am Torre, and I do not know how long ago it is since I wrote a single line.

xii:1 'Die Vorlage der Würzburger Glossen kann unbedenklich ins 7. Jahrh. datiert werden.'--Rev. Celt. vi. p. 319.

xv:1 See Rev. Celt. vi., pp. 322 and 328.

xvi:1 Prof. Zimmer also claims our text for this century. His words are (l.c., p. 261): 'Der Text gehört zum ältesten was uns von irischer profanlitteratur erhalten ist: seine sprache ist sicher so alt wie die ältesten altirischen glosses; er kann also noch dem 7. jh. angehören.'

xvi:2 With reference to Prof. Zimmer's well-known theory as to the Norse origin of Ir. fían and its derivatives, I may mention that the word fénnid occurs in 56.

Next: The Voyage of Bran