# Irish Lexicography

TODD LECTURE SERIES.

VOL. II.— PART I.

IRISH LEXICOGRAPHY:

An Introductory Lecture.

BY

ROBERT ATKINSON, M.A., LL.D.,

Professor of Sanscrit and Comparative Philology in the University of Dublin;

and

Royal Irish Academy's Todd Professor of the Celtic Languages.

DUBLIN:

AT THE ACADEMY HOUSE, 19 DAWSON STREET

SOLD ALSO BY

HODGES, FIGGIS, & CO., GRAFTON-ST.

And by WILLIAMS & NORGATE,

 London: 14, Henrietta-street, Covent Garden Edinburgh: 20, South Fredrick-street

1885.

PUBLICATIONS OF THE ROYAL IRISH ACADEMY.

TRANSACTIONS: Quarto, boards; or in Parts, stitched.

Vols. I. to XXIII. (some out of print). For Prices, &c., inquire at the Academy.

 Vol. XXIV.— Science (complete), 31s. 6d. Polite Literature (complete), 5s. ${\displaystyle \scriptstyle {\left.{\begin{matrix}\ \end{matrix}}\right\}\,}}$ together(with title) 24s. Antiquities (complete), 19s, 6d.

Vol. XXV. — Science (complete), 37s, 6d. Vol. XXVI.— Science (complete), 35s.

Vol. XXVII. (Polite Lit. and Antiquities) : —

Part 4. — Fasciculus of Prints from Photographs of Casts of Ogham Inscriptions. By Sir Samuel Ferguson, LL.D. (Witn Plates I*, to V.) [June, 1881.]

Part 5. — On Sepulchral Cellæ. By Sir Samuel Ferguson, LL.D. [November, 1882.]

Vol. XXVIII. (Science) :—

Part 12. — On the Embryology of the Mammalian Muscular System. No. I. — The Short Muscles of the Human Hand. By Bertram C. A. Windle, A.B., M.B., B.Ch., and Senior Moderator in Natural Science, University of Dublin; Pathologist to the General Hospital, Birmingham. (With Plates III. and IV.) [January, 1883.]

Part 13. — Report on the Acanthology of the Desmosticha. (Part II.) By H. W. Mackintosh, M.A., M.B.I.A., Professor of Zoology, and Curator of the Museum of Anatomy and Zoology in the University of Dublin. (With Plates V. to VIII.) [June, 1883.]

Part 14. — Report on the Acanthology of the Desmosticha. (Part III.) Further observations on the Acanthology of the Diadematidæ. By H. W. Mackintosh, M.A., M.R.I.A., Professor of Zoology , and Curator of the Museum of Anatomy and Zoology in the University of Dublin. (With Plates IX. and X.) [July, 1883.]

Part 15. — Expansion of Elliptic Functions. By William Nicolls, B. A. [April, 1884.]

Part 16. — On a Geometrical Transformation of the Fourth Order, in Space of Three Dimensions, the Inverse Transformation being of the Sixth Order. By L. Cremona, LL.D. Edinb., F.R.S., Professor of Higher Mathematics in the University of Rome. [July, 1884.]

Part 17. — Catalogue of the Earthquakes having occurred in Great Britain and Ireland during Historical Times; arranged relatively to Localities and Frequency of Occurrence, to serve as a Basis for an Earthquake Map of the Three Kingdoms. By Joseph P. O'Reilly, C.E., Professor of Mining and Mineralogy, Royal College of Science, Ireland. (Plate XL) [September, 1884.]

[For continuation of List of Publications, seepage iii. of this Cover.]

TODD LECTURE SERIES.

VOL. II.— PART I.

IRISH LEXICOGRAPHY:

An Introductory Lecture.

BY

ROBERT ATKINSON, M.A., LL.D.,

Professor of Sanscrit and Comparative Philology in the University of Dublin;

and

Royal Irish Academy's Todd Professor of the Celtic Languages.

DUBLIN:

AT THE ACADEMY HOUSE, 19 DAWSON STREET

SOLD ALSO BY

HODGES, FIGGIS, & CO., GRAFTON-ST.

And by WILLIAMS & NORGATE,

 London: 14, Henrietta-street, Covent Garden Edinburgh: 20, South Fredrick-street

1885.

DUBLIN:
PRINTED AT THE UNIVERSITY PRESS,
BY PONSONBY AND WELDRICK.

IRISH LEXICOGRAPHY:

AN INTRODUCTORY LECTURE.[1]

When I first heard of the proposition to found a Todd Lectureship of the Celtic Languages in honour of that eminent scholar, I felt that there could have been paid to his memory no tribute that was more in harmony with his thoughts when living and working amongst us. It is not necessary for me now or here to dwell on the character of his work, or on his earnest zeal in behalf of Irish studies. To us of the younger generation his name carries with it a charm that is full of regret; the older generation, in this case rightly "laudator temporis acti," still dwells on the memories of the past with a sadness that is tempered by the thought that in our Academy we have at least done what lay in our power to shield an honoured name from the "scythe and crooked knife" of time.

I am sure that I am but speaking the sentiments of all who hear me in saying that Dr. Todd loved deeply, as he prosecuted zealously, the study of Ireland's past history. It is not needful that we should agree with all he said, or accept all his theories, to enable us to value his labours in antiquarian research: he had a true feeling of the worth of this order of studies, a keen perception of the class of facts to which prominence is to be given, and a subtle skill in arranging artistically his collected material—an extremely rare combination which, in my opinion, belongs to only one living man amongst us—the author of the "Life of St. Columba."

Dr. Todd was not a linguist in the narrow sense of the word; he had studied, and carefully studied, the monuments of the ancient language of Ireland, but it was more particularly with a view to the elucidation of the history than of the language of the past. But he would have been the first to recognize the supreme importance, for that end, of the minutest investigation of the Irish language in its various epochs, and his swift intellect would have instantaneously grasped the bearing, and acknowledged the utility of the study of Irish in connexion with its sister languages, as well as with the wider family of which they are members. It was with this view that this Professorship was founded, and with no name could it be more appropriately connected than with the name of James Henthorn Todd.

The position, indeed, of these Celtic languages is sufficiently remarkable: the old Hindu records of the far East have been made to yield up their contents; their secret has been forced from the mystic arrow-heads that tell of the achievements of Oriental despots; and now, as one after another the members of the great Indo-Germanic family have entered the magic circle of linguistic, in which their hidden essence is wiled from them, so it has come that the westernmost of them all has fallen under the enchantment, and is waiting for the powerful voice of the magician to evoke the long silent soul.

There are phenomena in these Celtic languages that make them the delight and the despair of the linguist. The derivatives exhibit an abundant growth that speaks of a luxuriant soil, but the roots are covered deep down beneath layers of structure presenting a sturdy resistance to excavation—the stem breaks sooner than yield its hidden origin.

The known grammar of Irish is full of the strangest and most perplexing puzzles: thus, when we look at the compound preposition

 lium,﻿ linn, leat, libh, leis, leo,

we are involuntarily hurried into a different field of languages: these combinations, one fancies, might be Hebrew or Hungarian, Tibetan or Tamil! No wonder that in the early stages of the study a Yallancey or a Betham connected the ancient Irish language with the languages of the East, and gravely compared Celtic roots with Chinese or Phenician vocables. But the days of pardonable ignorance are past, though occasionally a belated philologer emerges from the depths of a thorny linguistic forest with a proud shout of "Eureka" as he holds up to view his long-cherished whims.

Criticism is more severe now that definite progress has been made, and perhaps some of the charm has been taken away by the accumulation of positive knowledge. Still there is abundant scope for discovery; for even since the foundation of this Todd Professorship a rich treasure has been recovered from the depths of the long past: the resolute student has successfully wrestled with the stubborn sphinx. Last year I delivered a lecture in Trinity College on the nature of a law, to which at the time I gave the name of "Zimmer's Law," in honour of its discoverer. [I did not, and could not, know that there was in print at the time, though unpublished, the narrative of substantially the same investigation by M. Thurneysen, with whom, therefore, apparently, Professor Zimmer will have to share the honour of discovery.] The importance of this 'find' can hardly be overrated: the study has been renewed with fresh zeal, and I trust that a generous rivalry will animate the leaders of this study to further efforts and yet more successful issues.

In times not so long past there were in Ireland societies that vied with each other in publishing Irish texts, but they have yielded now, as it would seem, to the all-devouring exigencies or desires of the present; and the modern societies, in urging the study of Irish as a vernacular tongue, have not had the slightest result in the way of increasing the number of students willing to devote themselves to the requisite training for the further prosecution of their studies in the right direction.

It is true something may be said also in their behalf; for, it may be asked, What facilities are afforded them for taking the right course? The Academy has indeed placed in the hands of the public the excellent facsimiles, far plainer to read than the original MSS.; but how is the Irish student to get at them, and how shall he procure for himself the key to unlock the wicket-gate at the entrance? That is a side in which the Academy is bound, I think, to give some help, and therefore, during my tenure of office as the Academy's Todd Professor, I shall endeavour not to leave out of sight the wants of the class whose sympathies and assistance I would seek to enlist. The field then is open, the labour is abundant. I propose to consider in this Paper, What are the lexicographical means at the command of the student who desires to take part in the work?

The grammar is fairly understood, but the dictionary is as yet not in existence; and each student will have for some time to make his own dictionary out of already published matter, until the time arrives for the birth of a Celtic Thesaurus.

The natural source of the modern vernacular goes only a slight way towards the goal; we must have recourse to the written memorials; we must go back to the earliest times of the existence of this family of speech, and trace its course through all the points of its orbit.

I. In speaking of the sources whence our knowledge of Irish is derived, we naturally turn to the earliest records, the inscriptions in the Old Gaulish language. These inscriptions, which unluckily are not numerous, are besides of extremely uncertain rendering. There is hardly a single word read in these monuments concerning which doubts are not legitimately felt. But, from the very nature of the case, perhaps all that these inscriptions will ever avail to us is, that the case-endings which are recognizable correspond fairly with the inferences made from other sources as to the prehistoric case-endings in Old Irish.

Quite as much difficulty awaits the student in the case of the Irish stone-records, commonly known as ogam-stones, which in any case are not likely to contribute important knowledge, either historical or linguistic. It is not easy to see at what point of Ireland's history these inscriptions could have been made. Prom the time of the introduction of Christianity, to the period of the Zeussian glosses of one thousand years ago, there must have been a considerable literary activity in Ireland, and the young, flourishing Church, was very unlikely to neglect or corrupt the language with which they sought to persuade and control the people. The very fact that these clerics studied carefully the Latin language would have made them the more desirous to preserve forms of their mother-tongue so extraordinarily like those of Latin as some of these ogamic inscriptions seem to possess. And besides, it is not to be left out of consideration, that even the early Zeussian codices are only fragments of a continuous activity, and that therefore they would have preserved relics of that older [assumed ogamic] state of the language, had any such existed, handed down from the first missionaries, who must have left Ireland at a time when the language was still in possession of these wonderful primeval case-endings. But nothing of the kind I refer to is found in these codices. The suggestions hitherto made appear to me to demand a rate of change which the circumstances of the case do not permit, and which nothing but a phonetic epidemic of a malignant type could possibly explain. But the course of linguistic transformation in a country at rest from serious troubles, and where, too, no external shock precipitates the progress, is far too slow to permit the assumption that such forms as qurimitirros and glasiconas [Ir. glas-chon] could be real living genitives at any period subsequent to the introduction of Christianity. Either, therefore, these inscriptions are much older than they are generally considered to be (which I do not think possible), or, we have not got at the right solution. For my own part, I do not believe that they represent the spoken language of the time when they were inscribed, but rather that they were a secret writing based on the language of long-past centuries kept up, among the druids and brehons, combined with other cryptic methods of writing that had no reference to the antecedent state of the language.

II. Leaving then this barren field of inscriptions, we advance to the sure and fruitful sources of the MSS. in Old Irish. The earliest of these are the so-called Zeussian codices, from which was drawn the material used by our great master in the construction of his vast work, the Grammatica Celtica.

[1853] The progress of study on these important texts will be best seen from a chronological statement of the works published subsequently to the appearance of that work in 1853. Zeuss made use of seven MSS. of varying extent, but whose language, according to the master, was "una eademque formis suis et regulis certis circumscripta, lingua hibernica vetusta" (Gr. Celt., p. xxxiv.). These were MSS. of the 8th and 9th centuries, from St. Gall, Würzburg, Milan, Carlsruhe, and Cambray, containing glosses and phrases explanatory of passages found in Latin MSS. of Priscian's grammar, or parts of the New Testament, and the Psalms.

[1866] The next step was the publication of Goidelica, by Whitley Stokes, in 1866, containing his transcripts of the Irish glosses found in MSS. at Turin, Milan, and Berne.

[1869] Three years after the publication of Stokes' book, Nigra gave to the world an edition of the Turin glosses, with a commentary on each word, and a considerable amount of explanatory detail.

[1871] Shortly after, Ebel, who had been steadily labouring in investigation of the structure of the language, on which he published many important papers in the Beiträge, was enabled, in 1871, to republish the Grammatica Celtica with many valuable alterations and additions;[1872]  and the following year Stokes re-edited his Goedelica, with much fresh matter, including glosses from Vienna, Nancy, Berne, Leyden, together with the Old Irish hymns and glosses found in the Liber Hymnorum, as well as the Irish notes in the Book of Armagh.

[1872] The same year, Nigra published his Reliquie Celtiche, giving a detailed account of the MS. of St. Gall, its form, pagination, contents, period and place of compilation, its ogam entries, its marginal readings, its script, with four illustrative photo-lithographic pages, and a selection from the glosses themselves.

[1878] The next important step was taken in 1878 by Professor Ascoli, who, in vol. v. of the Archivio Glottologico Italiano, began the publication of the Milan Codex, under the title Il Codice Irlandese dell' Ambrosiana, of which the second fasciculus was issued in 1882. Between the two fasciculi he published the St. Gall glosses in their entirety, with catchwords from the Latin text of Priscian, which, unfortunately, are not full enough to obviate the necessity of a constant and troublesome reference to the editions of the Latin text by Putsch or Hertz.

[1881] But now, in 1881, a most valuable addition was made to our means of study by the publication of Güterbock and Thurneysen's Indices Glossarum et Vocabulorum Hibernicorum quae in Gr. Celt, explanantur. This excellent piece of work enabled everybody to see at once whether any word was discussed in the Gr. Celt., and, at the same time, brought together the whole of the texts (so far as they are quoted in the Gr. Celt.) of the St. Gall, Turin, and Würzburg Codices, and of the Milan Codex as much as was then attainable; with singularly accurate reference to the place of occurrence of each word in the Gr. Celt., and a suitable classification of the different forms. This work furnished at once a convenient text-book and glossary of the Old Irish material, and freed the student from the enormous labour which had up to this time devolved upon the beginner, of making for himself some provisional index to Zeuss.

In the same year, a further step was taken by Professor Zimmer in the publication of his Glossae Hibernicae, a revised edition of all the Old Irish glosses, except those of St. Gall and Milan, which the edition of Ascoli had already made publici juris.

It is with very considerable eagerness that the remainder of the Milan Codex is awaited from the able hands of its editor, for it is not too much to say that every page of this work[2] adds some valuable item to our stock of knowledge.

Thus, when Ascoli's edition of the Milan Codex shall have been published, the student will have at his command about the entire body of Old Irish extant. It is therefore not too optimistic to expect that, before the completion of the present decade, we shall be in possession of a dictionary in which the whole of the Old Irish material will be sifted and arranged in a manner that will leave little scope for further amelioration.

III. The next important element is the glossaries. From this side, too, something has been done, but not much. The small glossary of O'Clery was edited by A. W. K. Miller, in the Rev. Celtique of 1879-80, tome iv., 351; but already in 1862 Stokes had published the more important glossary attributed to Cormac, together with the valuable law glossary of O'Davoren. In 1868, Stokes edited O'Donovan's translation of Cormac's Glossary; but the O'Davoren remained untranslated. The latter was submitted to a careful investigation by in an admirable Paper in the Revue Celtique, n. 453, and the publication of the Felire by Stokes, in 1880, gave an opportunity for the further utilization of O'Davoren, who still, however, remains untranslated.

In the Philological Society, 1859, Stokes published, also without translation, some glosses from MS. H. 3, 18, Trin. Coll., Dublin, and the same author in 1860 brought out a valuable little tract on Latin Declension, with Irish Glosses, accompanying it with his usual painstaking indices. But beyond these publications nothing of importance has been done in the way of facilitating the study, or at all events of completing the publication of the extant glosses. This is much to be regretted, because the glosses do undoubtedly contain much excellent matter, calculated to furnish invaluable assistance in the translation of our Middle Irish texts, as in many instances the statements of the glosses harmonize so well with the sense of the passage, that we can have little hesitation in assigning to them the alleged meaning. Thus, in Windisch's Irische Texte there are numerous passages where a knowledge of the glosses would have been of service; e. gr. on p. 280, dóit fri dóit leóit fri leóit, fuamain fri fuamain, gualaind fri gualaind, where leóit is explained in the glosses to mean elbow, and fuamain rendered by side, both words being apparently unknown to the editor. Again, p. 266, Windisch gives imbri bró, inserting the words in glossary under imbri, but without explanation. They are glossed bri = fearg, and bró = bruṫ, 'anger's glow'. Again, p. 260, the passage cíar bo mór trá a mainbech dorat im na mná aile, dorat a thri chomméit im Emir, is rendered by O'Curry [Manners, &c., iii 26], "so, though great the flattering praise he bestowed on the other women, he lavished (thrice) as much upon Emer", following the gloss, which explains muinbech by molaḋ; but Windisch prefers O'Reilly's explanation of sting, deceit, treachery, probably in ignorance of the gloss.

In many cases these glosses furnish information that may be accurate enough, but which will have to be taken on faith, because the word explained is not likely to be of common occurrence. Thus, when we meet the detailed account of a word, such as that given LL. 186 β 57, on the word srinci, viz. "nomen alicujus partis parvæ quæ sit in ore infantis in utero matris" (translated into Irish in O'Clery sub voce), there seems little likelihood of our getting beyond that statement. But then that is just the kind of word concerning which the tradition was likely to be kept long alive. Unfortunately, many of the glossarial entries are plainly of a speculative nature, only to be accepted by way of hypothesis till verification. O'Reilly's dictionary was compiled largely from old glossaries, and therefore is to be handled cautiously; but I am not inclined to denounce O'Reilly, who could hardly have been expected to do much better with the materials at his command; and I confess that I have at times found him quite correct in cases where I had doubted his information. Thus the word tioruirse 'remnant', 'fragment', was for some time marked in my copy with an obelisk, as possibly a vox nihili, till a wider acquaintance with the literature assured me that it was genuine, and used in that sense; e. gr. LB. 118 a 19, 'di-ambe din tiruarsi de', u if there be a remnant left of it "; cf. LB. 1 57 α 9 ni h-innister sund acht tiruarsi do na piannaib-sin, " only a remnant of the tortures are described here" ; 0' Donovan (Gr. p. 457) has an instance from F. Mast., a.d. 1174, and cf. the entries sub ann. 733, 1133, and O'Curry, Led., p. 555. Again, O'Reilly's entry of arada 'a severe punishment,' was to me a matter of doubt till I found the word referred to; it is a fem, n-stem, as may be seen from (sg. nom.) LB. 154 α 59, ba hí so aradu dobertha forru .1. a ngabail ar tus, a mbualad ocus a sroiglead co na facbatís cnáim na ball i cóir re cheli díb, "this was the torture that was inflicted on them, viz., to seize them first, and then to smite and scourge them till they did not leave a bone nor a limb undis jointed " ; accus. LB. 165 β 24, o 'tconnairc tra Iudás in tan-sin in aradain tuccad for Isu, "when Judas saw the punishment that was inflicted on Christ" ; 169 α 10, tabraid aradain cóir fors-in anmain i fil delb ocus cosmailius Dé, " inflict, ye just punishment on the soul", &c. ; but cf. also LB. 49 α 63, doronsat cuiccnecht in n-úain cháscda ocus a aradain, "they prepared the cooking and the punishment (?) of the paschal lamb".

The material in O'Reilly is of the most varied kind, but it covers an enormous area, and must be judged somewhat leniently. But even the excellent digest of Windisch, which only refers to a very small section of the literature, is by no means impeccable. I do not propose to go into a detailed examination, of course, nor have I the slightest intention of disparaging the work; but I give here a few examples to show that whether based on old glossaries, as O'Reilly's, or on extant translations, as Windisch's, all dictionaries for the present have to be regarded with suspicion.

Thus, in the Grammatica Celtica Zeuss had quoted as an example of a substantive ending in -ise, the word semis {gl. attenuatio), Ml. 22d; Windisch must have looked this up in the Ml. codex, for he cites the gloss, but he would not take the hint of the gloss itself, 22 d1, which has 'armúsemise' 'pro ipsa mei adtenuatione', i.e. mu sémi-se, with the part. augens. The word sémi [53b28] 'tenuitas' is common enough; but semise should be deleted as a non-existent form.

In his glossary, Windisch gives the very common word menmarc, 'darling', with only a (?) appended, though one thinks the meaning might have suggested itself even at first sight. It occurs in the Fled Bricrend, p. 260, § 18, at banlendan ocus at menmarc fer n-domain uli; p. 288, § 68, 13, is menmarc ban buaignigi.

The following passages will illustrate its use: — F. Mast. n. p. 1626, ó nách menmarc latsa gialladh do mac M., "as it is not thy wish to give hostages" ; ibid., gur an lucht bá menmarc lais do bheith ina fhochair, "with that portion of the army which he wished to accompany him " ; cf . F. Mast., m. p. 1706, 2016, &c. Windisch adds a singular conjecture of his own, asking {sub voce) if menmarc may not have arisen from menchomarc, giving the passage where he finds the word in an article by Stokes {Beitr. i. 340) :

gaidil, gaidil inmain ainxn
ise menchomarc a gairm :

thus translated : ' gaidil, beloved name ! my sole wish is to invoke it'; m'en-chomarc, ' mein einziger wunsch' : so that Windisch's suggestion amounts to this equation : —

menmarc = menchomarc = m'en chomarc.

I do not think any speculation of our native glossators could beat this.

Windisch renders bascaire "beating the hands together in lamentation " , quoting Stokes ; but the word has no necessary reference to sorrow. O'Dav. uses it to explain lam-comairt, l hand-clapping ', and both are found in LB. 141 α 3, 222 β 10, 224 α 27 ; «?/. also LB. 5 β 34, 154 β 35, 235 β 38, 259 α 39 ; F. Mast., m. 2292 ; but it is also used in the sense of clapping hands for jot/ ; cf . LB. 230 β 37, o atchonnairc Iudas sin, ro-s-gab for bascaire moir fri met na foelti, &c., " he took to clapping his hands from the abundance of his gladness ".

On p. 32 of the Texte, he has a remark that " a word comlabar [sic] in the meaning of ' speech ' has not yet been established ' ; and in his glossary he appends a (?) to the entry. But cf. Ml. 31b 84 : huare dK asne gnim tengad comlabrae is immaircide andurigni Duaid ingnimsin intengad duairbirt argnimaib inchoirp olehenae, " inasmuch then as speech is an action of the tongue, David was justified in employing the action of the tongue for the actions of the body in general": we could scarcely have a clearer example of the use of comlabra in the disputed ON IRISH LEXICOGRAPHY. 13 sense. Or, again, Ml. 3 Id 10 (amal ni cofil ni arachoat anargat nglan acht a techt inaicdi) sic comlabrai inchoimded iwfolngar gnfm disuidib fochetoir iarnalabrad, "just so the word of the Lord becomes act immediately on its utterance". Or cf. its use in Ml. 46a 15 : inchom- labrae, glossing (rationabilis) allocutio. It is clear, therefore, that Stokes was quite justified in translating comlabra by speech. The simple labor is brought into play in the Gr. Celt, in a com- pound which has nothing to do with it ; p. 3, foot, " eslabre (gl. ama- bilia), i. e. non superba, from adj. lobar superbus, arrogans". This is not correct, for the meaning is established as largeness (of heart) ; cf. Ml. 19c M , eslabrae, gl. dispensatrix Dei liberalitas; 57b 1 , gl. [manum] suae liber alitatis extendere ; 57b 6 , sua liber alitate sustentat. And the passage in Wb. 24b itself shows that no adjective is here intended, though the word is given as a gloss on i quaecumque amabilia ' [Phil, iv. 8], because the analogous clause ' quaecumque pudica ' is glossed 'buith cen pecca^T, '8inlessn,ess', so eslabre is employed in the constant meaning of liberality. O'Reilly has the word, viz., easlabhra,

• bounty*, ' courtesy', ' affability', and supplies the corrective in the

• from the great [deep, stubborn] silence.' And indeed O'Davoren

gives the word dedhoimn .1. demin no calma, with a fair appreciation of the meaning.

Again, Ascoli has edited a vox nihili in two places, which a fuller consideration readily sets aright ; Ml. 34 b 8 , the passage, ' qui devo- rant plebem meam sicut cibum panis ' is explained thus : " am nadn- gaib lius disuidiu issamlid insin nisgaib som lius difordiuc laimmi- muthaithese," where the word fordiue by itself is unmeaning : the word is di for diuclaimmim mu th[u]aithe-se, "from the swallowing of my people ' This word was evidently a difficulty to the learned editor, for on Ml. 19 d 5 , he gives it in the same fashion in the gloss on the passage, omnia ut fors tulerit aguntur incerta et more vivitur bestia- rum,' there is nobody who is guided by the dliged remdeicsen l dispen- satio rationis,' to distinguish between the good and the bad, 'do dechrugud etir maithi ⁊ ulcu'; then follows the gloss [:] ocad .i. a con- bad dliged remdeicsen oco-tuistin sidi acht inti bed tressu dofordiuc la : : : : alaile, "the one who is stronger devours the other," explained in the next note, ishe di ambes adi intf diib bes tresa oreaid alaile, where oreaid is the equivalent of the word in the former case, viz. do- fordiucla[id], to judge from the space. As to the form -aid, we have lenaid, Sg. 9 b 17 . But to divide dofordiuc is to construct a new word.

The same thing seems to me to have been done in an interesting case in the Irische Texte, where Windisch, in the Corrigenda, has adopted an emendation of Stokes, without apparently exercising the right of private judgment on the matter. He gives, " p. 559, read fo-glenim, 'I surpass' ; Stokes, cf. fod glein eminere, Ml. 37b." Now the passage referred to runs as follows, 37 b 12 : — sainred neulais leu isindan frisgniat sech each fodglein olehene, i.e. " they have a specialty of knowledge in the art they practice beyond everybody who learns it in general" (gl. quos multa peritia facit ab ejusdem artis consortibus eminere). But the eminere in this case refers to the word dor o scat, so that the suggested meaning of foglenim has no basis.

In many cases the words in these early texts are of rare occurrence in the later literature, though their meaning is tolerably certain from the context. I shall quote here a few instances of these rarer words.

erelca : Ml. 28 c 1 inerelcaib, gl. in insidiis ; 30 a s ba imfortgidiu ⁊ ba hitemul dugnith Saul conamuntair intleda ⁊ erelca fri Dd, " it was in murk and darkness that Saul with his folk laid plots and snares against David."

dindaitach: gl. reverens in Ml. 32 b 8 , 35 b 8 , in conjunction with ermitnech feid.

fwrgraid: Ml. 48 d 23 indfurgraid, gl. nihil reliqui; 57 d 8 innafur- grad sin, gl. reliquias ; 46 b 10 inraba cech n deithidin domundai huaim furgratae, gl. abjecta omni cura reliqua.

mesbaid: Ml. 19 c 16 ambandiuscartae [cf 32 d 16 ] inmesbaid .i. indebaie?, ' deposita simultate ; 50 c 18 simultate .1. homesbaid .1. de- baid.

debe : in the phrase ' debe tintuda ', ' difference of version ', of fre- quent occurrence in the Ml. codex, cf. 45 d 2 , 46 c 5 , 47 a 14 , c 16 , d 1 , 50 β 12 , c 2i , 53 c 6 , 54 c 8 , d 19 , 57 a 9 , 58 a 14 , &c., referring to passages in which the, text differs from the Vulgate. Thus Ml. 50 c 21 on the clause ' ego autem dixi in pavore meo' we have a ' debe tintuda,' ' ego autem dixi in excessu mentis' [Ps. 30, 23] ; so 50 b 12 , the text has * in manibus tuis tempora mea? but the Vulgate sortes mea, &c. And this explains Ml. 40 a 20 ata debe mec nand archiut forgnuso, gl. quibusdam verbis com- mwtatur, ' there is a little variation [m bee] with respect to the form.'

erassiget : Ml. 24 d 1 serassaigthe, gl. consilium Chuisi fuisset melius adprobatum ; 51 b 27 invitee spei esse omnia quibus praeter Dominum mortales exultant, glossed seraissaigthe .i. na chomairle dong[ni] duine sech dia noanesrassaigedar Dia, " any counsel which man makes with- out God, God annuls it" ; 34 d 8 qui ea (quae loquitur E.) irrita non esse patiatur, gl. beta n-serasaigthi ; 51 d 17 reprobrat, gl. ssrasaiged son. So that the passage from Wb. 30 b na herassiget dognfma (anasbere), is rather i that thy deeds may not make null what thou say est,' than ' ne contradicant facta facta tua' of Ghr. Celt.*, p. 444, referring to the operarium mconfussibilem of the text. emeeh: Ml. 27a s , isnaib emechaib no honaib fortrommaib, gl. in opportunitatibus in tribulatione ; 48c 9 , indemech, gl. opportune ; 50b 7 , cen emigi no fortrummai, gl. absque opportunitate ; cf. the following passage, M. Rath, p. 14 : —

is tu dlíghes co h-eimeach
at fisigh, at fírchléirech,

• This periphrasis of the infinitive mood of verbs used in connexion with finite

• I do not know why "Windisch has added to Crowe's failings of translation in

this case. Crowe was publishing the text of LB., and he edited senite from that text, with the translation of nets, for which no doubt he held Cormac' s sen, bird-net, as sufficient warrant (" sen .i. lin a ngabar eoin", p. 41, Gloss. Cor.), but Windisch, sub voce ti, quotes the Lib. Hymn, text of Stokes (in which the LB. senite is side tkt), and then stigmatizes with a ! Crowe's rendering nets, as if Crowe had read ti. No doubt the lib. Hymn, text is the better, but Crowe was render* ing the LB. text, and should not have been scourged here. of the vengeance of God on the people"; cf. LB. 153 α 12, droch- celmaine ; 14 mi-chelmaine. So that its meaning above is clear : "it seems to me that they are performing divination of wizards, for they take no single step without looking up, and they are arguing and conferring with each other". Windisch inserts saingnusta in his glossary from O'Davoren, where we have [p. 114] iar setaib saingnusta = iar conairibh sunnradachaibh. It is not uncommon in our MSS., cf. LB. 54 β 48, techtaid din in liachta-su etergna saingnusta as imchubaid fri crist ⁊ fri cech n-oen duine foirbthe is-ind eclais, " haec lectio habet et spiritualem sen- sum", a special meaning referring to Christ, &c; LB. 176 α 7, is d' oig is co saingnusta ro-genair mac De, " there is a special fitness in Christ's birth from a virgin". San-chan is entered in W.'s Glossary, from O'Donovan's Gram., p. 269, with the meaning hin undher, " to and fro " ; but it is somewhat wider than this ; cf. LL. 44 α 44, in nallach ro-s-marbsatar sain-chan im Lin ligda; "the folk whom they slew everywhere round the pleasant Liffey" ; Three Frag., p. 228, ro thionoil an rioghan iaram slógh mór impe san cán, " she collected from every direction" as O'Don. says ; LB. 52 β 54, tarrustar ann din na h-Iudaide sainchan im Ierusalem im-a-cuairt, Jews from all parts ; LB. 55 α 42, na, fer- anna examla sainchan in [leg. im] Ierusalem, " regiones circa Ierusa- lem undique"; cf. Nennius, p. 198, ted muir tar na cairgib moraib na muirbeach impi san chan, which Dr. Todd renders, " notwithstanding that the tide rises over the large rocks on the beach around it to and fro". Doraith has been translated with a (?) by Stokes in Irische Texte, 2 te . Ser., p. 9, 1. 197, "first of all." The following instances make the meaning clearer : LB. 38 β 21, ardaig . . . na ru-b guasacht bais do'n duine mine tesctar doraith o'n churp in ball-sin in ro-gein in galar, " lest it be fatal if the diseased limb be not cut out immediately "; LB. 250 α 27, ro-fiugrad din in ' pater ' hi secht senmannaib ro- sheindset na sacairt tall i n-Ericco dia torcratar doraith secht muir na cathrach, "the pater is foreshadowed in the seven trumpet-blasts they blew at Jericho, when the seven walls fell at once". The word JSsraiss is used LB. 5 a 35, atcondcatar . . . na ra-bi esraiss uilc do denam aice, " when they saw thatj it had no means of doing evil"; LB. 129 a 47, ro-suideged longport lanmor leis for muigib mor-r6di, ocus for esrassaib imaidble in dithrib, ocus for conairib comlethna coimeta ocus fethmi na coerech thret, where the meaning is explained by the similar usage in F. Mast, in. p. 2212, ní ro fhaccaibh conair na eassrus éludha, "he left no tohHot passage " [means of escape] ; ibid. p. 1896, ro bhaoí occ sccrudadh ina menmain do gres caidhe an tearsus eludha fóghébhadh), "he was constantly revolving in his mind the manner in which he would make his escape " [would get his means of escape].

Windisch gives Merken, Bemerken, as the meaning of foimtiu, no doubt following Stokes, whom he quotes on this passage, nathir arthuailchi ⁊ trebairi frifomtin cecha hamuis [Mid. Ir. Horn. p. 42], which is rendered by Stokes "a serpent in cunning and prudence for observing every attack ". But this is hardly the meaning ; cf. Ml. 43 a 16 do foimtin (gl. ad cavendum). It was in common use in Middle Irish, with varied application : thus 0' Curry, Zed. p. 641, translates a passage from LL. as follows: — " The doctor said to Conchobar that he should be cautious (co mbeth ifomtin), i. e. that he should not allow his anger to come upon him". A few examples from the P. Mast, will illustrate its employment : —

Yol. in. p. 1766, robdar cairdi a(n) eccraitte friss gó sin co na baoí hi foimhdin coccadh no comhfuachadh, " his friends were till then at strife with him, so that he was not prepared for war or hos- tilities"; ibid. 1980, ro líonta iaramh do laochraidh . . . i ffoimhdin toghla an bhaile, "they were afterwards filled with heroes," &c., "for the purpose of razing the castle"; ibid.2124, baoí ina foimhdin ⁊ ina fhoichill fri ré dá mhíos, " he was in wait and in readiness for him for a period of two months"; ibid. 2266, the same combination is rendered "watching and restraining the movements of the English"; ibid. 1988, ro baoí bheós i ttuile a aoise fri foimhdin imnigh ⁊ ettualaing an choccaidh i mbaoí siumh, " he was in the bloom of youth, and able to endure the hardships and toils of the war in which they were engaged"; ibid. 2252, pudar baoi leó fri foimhdin an choccaidh, " the powder which they had for carrying on the war".

The word amainse is defined in LB. 197 α 22, where it is used as a gloss in explanation of the names of one of the virtues : — 1. fides, in ires ; 2. castitas, in genus ; 3. humilitas, in umaloit : 4. caritas, in dercc; 5. prudentia, in trebaire .i. amainse ; 6. temperantia, in mesar- dhacht; 7. fortitudo, in t-shonarti; 8. justitia, in fhirinde; cf also 110 β 15, 118 α 25, &c. ; LIJ. 15 β 23, nir bo mor amainsi caich dib fri araili; F. Mast, sub anno 1086, liubhra lána d'amhainsi ⁊ d'inntleacht, " books replete with. genius and intellect; F. Mast. in. 2374, tigherna tend tóthachtach go ngaóir, go ngliocas ⁊ go n-amhainsi indtleachta ⁊ aigneadh, "a powerful mighty lord, with wisdom, and subtlety, and profundity of mind and intellect"; M. Rath, p. 148, it is rendered foresight, and O'Curry (Led., p. 580 penult, line) translates it cunning. But in LB. 118 α 25 this sense will not fit the passage — conad iar-sin dorat Dia nax plaga amaindsi irdarca ar Forand con ill-tuathib ilarda Egepti ar-aen fris; and indeed O'Donovan, in M. Ragh, 202, renders ina ma h-amainsi ocus na h-ainigne tucais ar Ultaib " [more becoming] than to have annoyed and insulted the TJltonians", &c.

The adj. amainsech is used in M. Rath, p. 160, ogh-briathra ána amhainsecha na n-airdrig, "the pure, noble, sapient words of monarchs"; whence the abstract in LL. 9 α 2, oc foglaim druidechta, ocus fessa, ocus fastini, ocus amainsechta, " learning wizardry, and knowledge, and prophecy, and amainsecht."

The word aprisce is of not infrequent occurrence in the LB., and its meaning may be fairly gathered from the following examples : — LB. 49 β 32, ro-fhetatar a n-enirte ocus a n-aprisce fen, " they knew their weakness and fragility"; 51 β 6, aimserind nw-fhiadnaise i ndlegar da each iressach etarscarud fri h-aprisce a thol collaide, " when it be- hoved each believer to separate from the inertness of his fleshly lusts " ; 39 α 4, is ercradach ocus is aprisc, used to gloss l caduca et fragilis '; ef. 193 β 33, ro-fhetar-sa at-aiprisce na doine, "I know that men are liable to fragility"; and 164 α 27, ro-thoirmisc umpu cotlud aimsire ocus utmaille menman, na ro-epletfs i n-aprisce peccaid, "he corrected (hindered) in them temporal sleepiness and instability of mind, lest they should die in the sluggishness of sin"; 165 α 61, co fesed Petar indus bud coir aircisecht do'n foirind dogentais imarbus tri-a aiprisce is-in eclais, " that Peter might know how it was fitting to commiserate the people who committed sin through inadvertence".

The following entry in "Windisch's glossary is unsatisfactory : — "direccra, p. 191, 18 : vgl. difhreagra, unanswerable, O'R." There is great virtue in a vgl. ; but I would rather "Windisch had given his opinion. Anyhow, O'R. is not to blame, for he did not assign this meaning to the word direccra. It is common enough in Middle Irish, and variously used as an intensive adjective. I [shall quote a few instances of this varied application from Z. Breac : —

(a) heavy, strong, of perfume [LB. 35 β 7] : in tan tucsam corp Stephain as in inad when we took the body of Stephen i mboi, dorala talam-chumscugud mor from the place where it was, there hap- arm, ⁊ tanic bolad direccra de as in pened a great earthquake, and there adnocul, co ro-lfn in uli inud i mba- came a heavy perfume from it out of mar : ro-b ailgen tra in mbolad-sin. the grave, so that it filled the whole place where we were — pleasant in sooth was that perfume. (b) heavy y loud, anguished, of screaming [LB. 39 β 52] : — is ann-sin nach fil comdidnad no cum- there is no consolation nor rest nor sanud na etarfuarad doib-sium, acht coolness for them, but heavy shouting, nuall tromm ⁊ diucaire dermair ra-mor and mighty, vast, loud outcry, ro-direccra. (c) heavy, thick, of darkness [LB. 118 α 27] : — in cet plaig tucad for in tir-sin .i. the first plague that was brought on dorchatu dfreccra. this land, viz. thick darkness. Cf. also 154 β 25, 41 ; 165 α 37, &c.

In his edition of Tochmarc Etaine, Ir, Texte, p. 129, we have a passage whose explanation escaped Windisch, viz. : " rotirmaiss ecaine ocus mor olcc ocus imniuth duit bith i n-ingnaiss do mna". In his glossary the first word is divided, and placed under (ro) tirmaiss, but no meaning is attached to it. I believe it to be ro~t-irmais\$, " hath hit thee", the word appearing under the forms ermaiss, urmaiss, and, as here, irmaiss [cf. forms like aurlam, urlam, erlam, irlam].

In Cormac's Gloss., sub voce, 'taurthaiV (' random shot'), we have urchar . ... do urmaise secip nach raeta, &c., which O'Donovan had rendered " a throw .... to hit anything whatsoever." Upon this Stokes remarks : — " I rather think this (urmaise) means ' to aim at', and then 'to purpose* ", quoting tuisled ho ermaissin firinne from Z 2 1064. But surely the very quotation makes for 0' Donovan's transla- tion. The full gloss is [Ml. 2d 6 ] is f uasnad dutmenmainsiu tuisled ho- ermaissiu firinne trimrechtrad natintathach, i. e. " it is a disturbance to thy mind, thy failing to hit the truth through the variety of inter- preters", aiebasenim te magis interpret urn varietate turbari. But, in order to make this clearer, I add a few more instances of its use : LL. 125 α 14, no-shined a shuanemain etar da chualli, ⁊ no-bid oc dfburgud eturro, ⁊ ni anad co n-ermaised in ubull no-bid for cind in chualli, "be stretched his rope between two poles, and used to shoot (sling stones) between them ; and he did not stop till he hit the apple that was on the top of the pole". From this meaning comes the fur- ther application, ' to attain', 'consequi', 'to have time for* : cf. the following passage [LB. 49 α 28] : —

b£s imorro 7 crabud in arain neiu- the custom and religious practice of descda, is as t&sech ro-tindscanad : d'n unleavened bread originated from the chet-chaisc ro-chelebairset meic Israel first passover that the children of Israel in Ramessi i tir Gessen is-ind 'Egipt, dia kept in R. in the land of Goshen, in tardsat forru na h-Egeptacdha co forec- Egypt, when the Egyptians compelled nech facbail in tire, co na ro-ermaisetar them to quit the land, so that they had fris-in tindenus descaid do cumasc ar not even time, from the precipitancy (of in mein amal ba gn£th ddib remi do their start) to mix the leaven in the flour, d£nam. as was their wont before to do.

This word seems to have been a constant stumbling-block to O'Donovan in his translation of Magh Rath, e. gr. p. 268 : —

ni bi duine ar doman gan a fhod urdalta airchennta oidhedha d'urmaisi, gin go raibe tacha tapaid ná esbaide engnama air, " there is not a man in the world for whom his certain and fixed place of death is not preordained, even though he should have no want of vigour, or lack of valour " (O'Don.). Here even the logical connexion shows that the sense is " who does not [hit, attain] meet with his appointed doom, in spite of his prowess", &c. Again, ibid. p. 192 : — ro pad feidm ocus ro pad urmaisi airig no fhír-laich fuirech re fégad a féinned ocus re taidbred a tuaruscbala, "it would be the business and improvement of a chief or true hero to remain to view these heroes and conceive their description " (O'Don.). In these vague, alliterative phrases, a certain liberty must be allowed ; but I do not think that the translator has hit the point ; and the true rendering seems to me something like this : "it would be a great etfort on the part of, and would only be attained by, a great chief to stay cool, inspecting his troops and reviewing their qualities " ; because, as the narrator goes on, with considerable insight into the native character, "these heroes are not mild to be commanded ; and it is a torment to chieftains to be re- strained 1 '. Two good examples oi this use of the word ' to happen to- gether/ of two events synchronizing, hitting the same point, &c., occur, ibid., p. 100, in adaigh ro h-urmaisedh ar Dhomnall do dirgudh ocus do oirdnedh i n-oirechus Erenn, as i sin adhaigh ro h-aentaighid na h-oirechta, "the night on which it was resolved that Domhnall should rule, and be elected to the sovereignty of Erin, was the night on which the assemblies were united," &c. This should be "the night on which it happened to Domnall, to be elected was the very night on which" , &c. This is shown better perhaps in the following passage, p. 106: do h-urmaised sén saerighda, soineamhail, do'n ardfhlaith ocus d' Erinn i comhrach re cheile, "the noble, happy prosperity of this monarch and of Erin were ordained together" (O'Don.) ; but the meaning is simply that the two events concurred. Similarly in F. Mast, we find the term used vaguely, with no due appreciation of the right meaning, e. gr. in., p. 2282, do rala fordal conaire [" wandering from the way" : cf. in., p. 2198] ocus sechrán slicchidh do na slocchaibh lá dobhar dhorcha na h-oidhche co nár urmaissettar a neolaigh saighidh gus an ionadh chinnte, "the forces mistook their road and lost their way in consequence of the great darkness of the night, so that their guides were not able to make their way to the appointed place"; n., p. 1452, ni ruacht las an sluagh ngaoidhealach dol in inneall nó a norducchadh amhail ro bha dír dóibh, ⁊ ní mó ro urmaissiot comhairle a naireach do ghabhail, "the Irish army were not able to go into order or array as was meet for them, nor did they take the advice of their chiefs " (O'Don.). In both instances the meaning is fairly enough given, but the force of urmais in the sense of 'hitting', 'falling in with', 'happening upon', &c, is not duly recognized and expressed.

It is, no doubt, unsafe to deduce conclusions from the etymological connexions of a word, but there are some words so peculiarly formed that they almost inevitably call attention to their origin. Thus a familiarity with the words etargne, etar-cne, 'cognition' (cf. omietarcnu,

• experiment ', Ml. 19 a", 27 a 6 ; cf. 19 d w , 42 b 13 ' 27 , etarcnaib,

etarcnu, etarcnae) ; and bis, ' custom ', suggests the explanation of beasgna, 'law', as bes-cna, 'the knowledge of customs ' : the early laws, are just the immemorial customs of the people, their manners, and hahits ; cf. Or. Celt. 986, and of Ml. 14 c 11 , is reid foglaim inbesgnai, "vitae ratio ad intelligendum prona." This word b4* enters into another compound word of rather curious import, viz., b&stindrim, used in Ml. to express Jerome's tropologia : vide 48 c 11 , trisin bestindrim .1. * aliud sonans, aliud sentiens'; cf. 41a 8 , trisin mbffistindrim, where bis corresponds to τρόπος.

Another compound of this adb is met with in adhbhchlas, glossed in O'Clery by aoibhneas, a meaning that somewhat disguises its origin. The second element is a derivative of the root clu (inclytus, κλέος, &c). It occurs in Ml. 40 d 19 , do adbchlois, as gloss on pompae ; and cf. ibid. 11 , amal bid hualailiu chlausul adbchlostu .i. trop, ' velut in clausula pompatica In middle Irish it is of common occurrence ; cf LB. 36 α 7, Ecclesiastes tra, lebor e side i fhollsigther diamine ⁊ erchra in tsaegail i n-a gloir ⁊ i n-a ddbchlos, " the vanity and perish-

• Ml. 36 a 10 rucle" .i. erdarcai, as a gloss on [οἱ] conspicui. There is nothing new

as to the meaning of the word, but it gives rise to a query. In the Gr. Celt., p. 164, it is explained as a compound of the intensive particle ro and gle, but why this composite should involve the change of tenuis to media it is not easy to see. Now, in Ml. 37 d 18 , we find 'isrw^solus, as a gloss on praeelara, and 37 d 8 , isnaih rt^etrachtaib, ' in praeclaris ; * and this suggests rug-gle as the origin of our rucli. ableness of the world in its glory and pomp " ; 45 α 36, tuarcaib do ind-sin in uli fhlathiusa in domain co n-a ngloir ⁊ co n-a n-adbchlos, "he brought before him [ostendit ei] all the kingdoms of the world, with their glory and their pomp " ; cf. 46 β 45 ; ibid. 235 α 42 ba mor tra diumas ocus ddclos ocus bocasach in rig cholaig sin, " great was the pride, and pomp, and arrogance of that profane king" [for bocasach, cf. 154 β 19, 256 β 33, lucht in bocasaig ⁊ in diumais]. In the F. Mast, it is found as an adjective, adbclosaeh, with the mean- ing renowned, cf. in. 2178. But I have never met it in the sense given by O'Clery, and O'D., Suppl., as pleasure , joy ; • and I do not think that meaning can be justified.

In some cases a word is wrongly rendered which, from its very constancy of meaning in numerous occasions of occurrence, might have been expected to be free of variation. But here something must be allowed for the persistent habit of alliteration. The speaker or writer has ever had in his head a vast number of alliterative vocables, which the very necessities of his training has made it incumbent on him to acquire, and which, once acquired, are constantly thrusting themselves forward. This use, of course, operates on the translator who, allowing himself a certain liberty on the ground of this tendency to accumulate otiose epithet, prefers to give in a general way the meaning of the sentence, rather than come to close quarters with the separate words. In the Atlantis, rv. p. 212, we have an example of the way in which, as 0' Curry says, "old Irish writers burdened their text with adjectives of intenseness".

The adjective ainiarmartach, which O'Curry renders unmerciful, occurs, F. Mast., m. p. 2288, bá huttmall anbsaidh ainiarmartach a ccomhairle, which O'Donovan translates "their counsel was hasty, unsteady, and precipitate" . The word means, as its analysis shows, in-consequential, for iarmart is of familiar use in the sense of consequence, though Windisch seems to doubt this in his glossary : cf . F.Mast., m. p. 1784, ro bhen a droch iarmairt don dúthaig, "his territory experienced the ill effects of it"; LB., 45 β 1, uair ro-fitir an iarmairt no-biad de iar-tain, " he knew the consequences that would proceed from it thereafter". Or again, M. Rath, p. 170, ni dat coimedaig inill iarmartach-su, "thou art not a vigi- lant keeper of a flock" (O'Don.), where indeed inill, 'safe', is translated as if it were indile, 'flock* ; LB. 166 α 55, olo tra an iarmairt ro-lecsit Iudaide f omi fen ann-sin .i. f uil Crist do thabach di-a channaib di-a n-eisi, " an evil consequence they brought on themselves in the exaction from their posterity of the penalty for the Blood of Christ " ; 0' Curry, Led., p. 395, gives " it will be a bad legacy to Erinn's land", from the original (p. 624), olc an iarmairt d'iath Eireann. But now, cf. M. Rath, p. 272, da ghuin ainmhíne ainiarmarthacha, " two fierce and terrible blows " ; cf. LB. 188 α 25, tocbaid a hum co tuc bulli aniarmartach ann-sin hi mullach chind a brathar, "he struck a fierce blow on his brother's head " ; F. Mast., n. p. 1 179 [ann. 1 170], gniomh anaithnidh ainiarmartach, "an unknown, atrocious deed".

I do not think the word ' tesmolta ' occurs in the F. Mast., but it is not infrequent in middle Irish, and never with the meaning to it attributed by O'D., who evidently deemed it a compound of tes, heat, and molad, praise, thus M. Rath, p. 106, gur ob do theasmoltaibh tigernais, "thus far the ardent praises of the reign of the mo- narch" (O'D.). But cf. LB. 36 α. 5, paraule sin lebor as-a forchan- ter in duine as a noidendacht i n-a besaib ocus imo'n tesmailt is coir do do shechem and do inntshamail, "Parables is the book out of which man is taught from his infancy as to his morals, and about the habits which it is right for him to follow and to imitate " ; LB. 183 β 52, nocon indraic do fhir m* oesi-sea (ol se) brecc no doilbiud do denam co ro-midet sochaide do na moeth-oclachu Elizar di-a n-ad slan nocha bliadan, do-thecht co bethaid ⁊ co tesmailt na hgSnti, " non enim aetati nostrae dignum est, inquit, fingere; ut multi adolescentium, arbitrantes Eleazarum nonaginta annorum transisse ad vitam alienige* narum" [Vlg.] " to thelife and habits of the gentiles" [II. Macchab. vi. 24] ; LB. 211 α 4, co n-id and-sin ruccad epistil uad do Dindim ri na Bragmanda co n-eicsed side do tesmolta a ndaine ⁊ a comairberta bith, " wherefore there was brought from him a letter to Dindim, king of the Brahmans, that he should tell him the habits of their people and their customs ".

But to return to F. Mast. — The very adjectives uttmall and anbsaidh, which in in. 2288 O'Donovan renders hasty and unsteady, are found in p. 2126 with quite a different (and wrong) meaning. A general divides his troops into two divisions, the one comprising his veteran troops, to maintain the fight, and the other, a ghille dhiana dheinmnedhacha ⁊ a occbhaidh utmall anbsaidh, which O'Donovan translates, " his swift and energetic youths and his nimble and athletic men" ; hut the contrast shows what is intended, which indeed is pre- cisely what the words mean, his unsteady , unreliable troops.

Sometimes the difference is more important. On F. Mast, iil p. 2272, O'Donovan has (unintentionally) distorted a historical fact hy a mistranslation: go ro comhairléicc an bhainrioghain ⁊ an chomhairle d'Iarla tuadhmumhan tocht go n-iomat long ⁊ laoidheng . . . do cabhair ⁊ do comhfhurtacht muintire an prionnsa i nErinn; translated thus : " until the Queen and Council advised the Earl of Thomond to go with many ships and vessels .... to relieve and succour the Sovereign's people in Ireland ". His note in loco says : " This is a very strange vert) to use. It should he ro fhorchongair or ro fhorail, ' requested or ordered' ". It is certain that any historian following O'Donovan's translation and note would altogether misconceive the Four Masters, for the verb means permitted, and has no reference either to advice or command. It is of so common occur- rence in this sense that I shall not quote any examples [of. Ml. 31c 14 , 32c*, 58c 6 , 38a 11 , 40d 6 , 44d 21 , 44d 16 > 20 , 54a 10 , 56c 7 , 57c 5 , 53d», &c.]. But O'Donovan has rendered the word wrongly in many places, in accordance with the misconception, e. gr. vol. in. p. 2050, " the general permitted [ro chedaigh] them to frequent Leinster . . . ., whereupon ro battar romh lás an ccomhairlecchad sin acc taistel ⁊ tathaighe gach tire ina ttimcell, "by this instruction they con- tinued traversing and frequenting every territory around them", instead of " through their permission ". Again, vol. 1. p. 178 [sub anno 1213] " [the steward] began to wrangle with the poet very much, although his lord had given him no instructions to do so ", gion gur bo hé a thiccerna ro chomhairleicc do.

One of the most fertile sources of mistake is the confusion of root syllables under the influence of the phonetic laws of the language; thus cur, * to put and gar, i to speak when compared with the preposi- tion frith, develop into the nominals/r^or andfrdcre respectively. The confusion of these and related forms has led to some curious renderings, e. gr. in O'Donovan's Suppl. to O'R. we have an entry frecor céill, 'an exact return', giving two quotations (from H. 3, 17, and H. 3, 18), and a reference to Zeuss 11. 1130. If we examine the use of the phrase in the Laws, we see that a connexion with the word fricre (responsum) is always at the bottom of the translation, and the sense is accordingly obscured : e. gr. Senehus Mor, n. p. 286, dligthur a fastad . . . . . muine frecurthur cell co foltuib techtaib, where the translation reads "it is right to make them bind- ing unless he (the chief) is responded to with lawful returns". In fact the gloss on the passage shows that the phrase had quite passed beyond the reach of the glossator, for he paraphrases muine freagartur é do rer incialla in breiteamuin, " unless the re- turn is made according to the sense of the judge ". Evidently here everybody looked on the words as meaning "response according to the sense". The second passage involves the same misconception (vol. iv. 98), ar ba headh ba techta leo frecor ceille in talmun dia torad fadesin, rendered "for they deemed it just that the land should receive (for the injury done it) an equivalent in its own pro- duce". But this is to introduce a subtlety foreign to the original gloss, which simply means "they deemed right the cultivation of the land for its own produce v . The use of the phrase is exemplified in the Gr. Celt. 9 , p. 917 ; cf. Ml. 30 d 16 , 43 a 2 .

The divergencies of the monachic and the brehon uses of the Irish language, as representing the Christian and pagan aspects of civiliza- tion, cannot, perhaps, yet be inquired into with much prospect of suc- cess; but it can scarcely be doubted that, though the laws were subjected to the early revision of the missionaries who converted Ireland, a work in which they were coated over with a varnish of ecclesiastical tinc- ture, the brehons kept the old institutions, on the whole, singularly free from subsequent infusions of similar tendency. The investiga- tion of the terms adopted by the monks will perhaps afford a clue t% the felt deficiencies of root- words expressive of the new ideas.

So many terms relative to books and writings are derived from Latin {cf. liter, focal, scribend, legend, aibgiter, caiptel, sillab, epistil, lebor, fers, serin, &c), that the inference seems unavoidable that no- thing of the kind was known in Ireland before the introduction of Christianity. The great number of words of ecclesiastical import is of course perfectly natural under the circumstances, though it is per- haps not unworthy of notice that the word for the Evangel is translated into the vernacular, so-scele, just as the word gospel. But in other words not of Latin origin, possibly traces of a differenti- ation of meaning may be discovered, depending on the changed cir- cumstances of the time; thus the word gor, 'pious'; cf ML. 44b 83 , donaib duthrachtib innan^or, 'votis piorum The negative ingor occurs (Ml. 56 b 9 ), imfolngi comrorcuin dosochaidi cidarabiat indfirien isnaib imbedaib ⁊ isnaib frithoircnib ⁊ indingoir isnaib imbedaib ⁊ isnaib soinmechaib, "it causes trouble to many why the righteous should be in tribulation and in sufferings, and the impious in wealth and prosperity"; ibid. 57 d 8 , med brithem- nachtae dee huandamnither intingor is huantfirinni inbrithemnachtae si conocaba infirian, " eadem libra judicii qua deprimitur impius adtollitur Justus". Thus in the Brehon Law glosses we find the two terms, gor and ingor, contrasted: cf. SM. (cf. also p. 52) n. 288, 1. 29, mac gor ocus mac ingor, " an obedient son, or a son who does not support his parent"; 11. 22.1. 29, cethraime eneclainne athar uil do mac ingor, " fourth of the honour-price of the father is [due] to the son who does not support his father". This rendering is taken from the gloss in H. 2. 15, given in O'Donovan, Suppl., maith cach macc bes gor di'athair, " good is every son who is ' pius' to his father." The words were in use in the ninth century as the equivalents of the Latin pius and impius ; were they specialized at a later period in the direction of Latin pietas, &c., piety shown in the support of parents ? { But the other theory is tenable : the earliest monks no doubt modi- fied the signification of many native words when applying them in reference to sacred topics, and it seems to me not improbable that they adopted these terms, gor and ingor, which really denoted ' behaviour towards parents', in the deeper sense of actions considered in their relations to God.

It may be objected that in many cases the difference is so slight that it is not worthy calling attention to such unimportant points ; and indeed Mr. Fitzgerald (in the Rev. Celtique of Oct. 1884, vi., p. 196) has denounced rather severely the whole of the modern school of Celtic students. I hope he will pardon me for suggesting that the ' arid treatment ' of which he complains is in reality more hopeful for the attainment of the knowledge which he himself desires. Much of the matter printed is, as he says, rubbish ; but it has one merit, that of containing words whose meaning can be fairly got at, thus enabling us to give precision and point to much that troubles students in obscurer texts. When Mr. Fitzgerald talks about Crowe's translation of the Amra Choluim Cille, we have to decide what 'translation' means. He may be quite sure that the interests of Celtic story, mythological, or other, are not being neglected during this long and arid process. Many a tough conflict of wordy criticism will have to be fought. A good deal of heat will have to be generated before the goal is attained. The battle has not been without wounds and clamour so far, nor is there any reason to suppose that Irish can be freer than other philology from the baleful results of literary vendetta. The truth is, that Celtic studies labour under the fatal disadvantage of having no competent public to oversee and control: the criticism is frequently harsher than the occasion deserves. But there is unmistakeable progress; enlightenment is spreading; the knowledge once possessed by the privileged few is now of comparatively easy attainment; and the leaven of an eager desire to get to the bottom of these Celtic mysteries is working in many minds. The difficulties are mainly lexicographical: we do not know the import of many a word that occurs in our Irish texts, and we are not permitted the free license of guessing indulged in by the past generation. Mr. Fitzgerald might remember that there is not a page of his article that does not contain speculations or references depending on the meaning of individual words. Take, e.gr., the following passage, p. 201: "This collocation or opposition of the Hound and the Dog reappears in the ancient division of Ireland into Conn's Half (the North), and Mog Nuadat's Half (the South): for Mog Nuadat is but the Servant-of-the-Hand … and there is reason to suspect that in effect the other name is the Dog's (or Wolf's) Half." Now this is a speculation based on the meaning of Nuadat and of Conn; and by what means can the settlement of the question be sought other than by the comparison of texts, in which these words are used in the meanings alleged? Words, words, words, that is what we want, and for many a long year they will be wanted. At present scarcely the simplest piece of Middle Irish prose of a few pages long but contains some word or phrase that must be passed over with a query or a blank space. It is useless arguing that Nuadat means hand, unless texts can be produced clearly establishing this meaning, which certainly has not been done yet. But it is astonishing how soon the note of interrogation drops away from a hypothesis.

My main efforts, therefore, will be directed to the publication of Irish texts of tolerably ascertainable meaning, so as to secure as large a number as possible of authoritatively established vocables, and at the same time to set forth in its true light the extant literature. Too much has been made, as I think, of the supposed relics of mythological lore contained in our Middle Irish MSS.: there is a good deal to be done on the simpler and more commonplace side, for there is abundance of material in these MSS. which it will repay the student to peruse, before he begins to expend labour in the wearying solution of poetical conundrum, or to dissipate energy over the wire-drawn speculations of the ingenious mythologist.

1. Delivered, April 13, 1885.
2. In the case of MSS. of this supreme importance, it is a matter of absolute necessity that every word, syllable, and letter should be scrutinized with the closest attention, so as to secure the exactest possible reproduction of these invaluable documents, which form the basis for all scientific study of Celtic speech. It is, therefore, perhaps to be regretted that the MS. could not have been photographed.

This work was published before January 1, 1923, and is in the public domain worldwide because the author died at least 100 years ago.