Auraicept na n-Éces

Introduction

The Handbook of the Learned, here edited for the first time, is a work that opens up many questions.

Éces is often equivalent to fili. Filidecht covered the whole field of poetry, romance, history, biography, geography, grammar, antiquities, and law. The poet-jurist, who, seated, gave judgments in verse, is probably referred to at lines 407,8. The Auraicept treats chiefly of the Ogham alphabet and grammar, but if the Trefhocul be included, it treats also of poetry in the strict sense.

The poets,filid, were a guild, making their own special laws,and exercising discipline upon their own members (2193). They claimed and used the right to quarter themselves and their retinue upon society (2221), and they exacted a fixed sum for their poetic compositions. In general this was cheerfully paid; the means for enforcing unwilling payment was satire. The exercise of this potent weapon was moderated by rule (1935), certain forms of satire, such as tamall n-aire (1932), being forbidden in the Trefhocul; and though the poets have been abolished by law for over a century, even at this day in certain districts the phrase, dheanamh aoir air, to satirise one, is not without its terrors.

The poets were a secret society with a language peculiar and intelligible to themselves only. According to their literary tradition Fenius, at their request, devised this language for them (195), and its obscurity was essential (21). The people often rose up against the poets and attempted to repudiate their claims. One such rising was that at Drumketta, A.D. 590 (1472). About that time they numbered 15,000. Owing to the advocacy of St Columba, himself a fili, they were suffered to continue, but under restrictions.

The filid were a strictly professional class, undergoing a rigorous training to fit them for their position. The bards, on the other hand, were unprofessional, and more or less untrained, but they practised a large number of metres in which the filid also were required to become proficient.

The following tables (cf. the later scheme in Joyce's Social Hist., i. 430), will show what place the Auraicept occupied in their studies.

The Fili, his Rank, Name, and Compositions, with the Rewards therefor, and his Retinue (2219–2254). Rank Name Metre Reward Retinue at feasts Ret. on circuit Ret. ordinarily Ret. at poetic feasts/contests I. ollam anamain a chariot (=one bondmaid) 24 8 12 10 II. anrad nath five cows 12 6 5 8 III. clí anair four cows 8 5 4 6 IV. cano emain one horse (=two cows) 6 3 2 4 V. doss láid one milch cow 4 2 1 3 VI. macfuirmid setrad one cow-in-calf 3 1 1 2 VII. foclóc dían one three-year-old heifer 1 1 1 1(?)

The Yearly Studies of the Fili. Each year included the studies of all preceding years. Year Name ^{Studies covered} 1 foclóc l. oghum, besides regular oghum; the Auraicept with its prologue and with its flexions; l. drécht, vi. dían. 2 macfuirmid l. oghum, besides usual oghum; vi. detailed lessons of filidecht; xxx. drécht; x. setrada, senamain, and snaithe senamna. 3 doss l. oghum, besides ebadach nIlmain; vi. other detailed lessons of filidecht; xl. drécht; xvi. láid. 4 cano l. drécht; l. bretha nemid; xx. emain. 5 clí lx. drécht; xxx. anair; xxx. iarmberla. 6 anrad lxx. drécht; lxxx. nath mór; lxxx. nath becc &ersir; berla na filed. 7 ollam brosnacha suad, i.e. the bard metres which the poet ought to know, for that is the poet's lesson of the seventh year; e.g. l. divisions of brosnacha, i.e. dechnad mór, and two species of nedchnad mór are there reckoned, viz. sned and trebrad. 8 ... fiscomarca filed .i. dúili berla &ersir; clethchor choem &ersir; reicne roscadach &ersir; láide .i. tenmláida &ersir; immas forosnai &ersir; dichetal do chennaib na tuaithe &ersir; dinshenchus, and all the principal tales of Ireland in order to relate them to the kings, lords, and gentlemen. For the fili is not yet perfect. 9, 10 ... xl. sennath .i. ; xv. luasca &ersir; vii. ena; eochraid of lx. words with metres and xiv. srotha and vi. dúili feda. 11 ... l. anamain mór l. anamain becc. 12 ... cxx. rochetal; iiii. cerda, i. e. cerd of Ladchend mic Bairchida, &ersir; cerd hi Chota &ersir; cerd hui Bicni, &ersir; cerd Béci.

A brief study of the Auraicept is sufficient to convince one that the leading extraneous source is the Latin Grammarians. Some of them are cited by name, Priscian (A.D. 450), Donatus (A.D. 350), Pompeius, and Consentius.

If it be urged that the quotations from these authors are a late addition to the Auraicept by way of learned illustration, it is answered that in any case the general setting of the matter follows closely the didactic style of the grammarians, as the following examples, occurring passim, will show:— Quaestio est, Gr. Lat. v. 537, 16, 29; 541, 20, 32. cest, Aur. 9, 57. Ouaesitum est, v. 228, 18, Quaeritur, v. 165, 27; 210, 38, De qua quaeritur, Origg, xvi. 10, 2 conagar, Aur. 1019, 1375. ut sciam, v. 195, 19. ut scias, v. 121, 15, 18; 173, 18: co fesear, Aur. 1577. ut sciamus, v. 10, 16. sciendum est, v. 180, 32: is soigti Aur. 3508, is fisid 3523. scire debemus, v. 277, 30. scire debes, v. 142, 15.

The matter itself of the Auraicept is largely identical with that treated of by the Latin Grammarians in their early chapters—the alphabet, classification of letters, sounds and syllables, consonant and vowel changes, gender and declension of nouns, comparison of adjectives, prepositions governing dative and accusative cases, the accent, artificial and natural, genus and species, and a few other incidental points. The omissions are almost equally significant. There is no classification of declensions, no declension of adjectives which are tacitly included with the substantives, no treatment of pronouns except as tokens of gender (aurlonn, 585), or as emphasised by féin = met (726), and the whole accidence of the verb is wanting. The similarity between Latin and Gaelic failed at this point. The paradigm of the verb is tentative and native (304, 653). An endeavour is made to show that, while there is a correspondence in meaning between the two languages, Gaelic is the more comprehensive (1081).

The language is Middle Irish, but the basis, which has been much worked over, all belongs to the Old Irish period.

The composition consists of Text and Commentary, the latter forming the great bulk of the work. The text is the oldest portion; the commentary, in parts as old as the text, was in a process of continuous growth. The text, written in a large hand in most MSS., is printed in leaded type. BB, here followed, curtails the text. The Book of Lecan and T. make a much larger delineation of text. The question as to what is text and what is commentary will require further study for a satisfactory solution, but it may be here remarked that much of the primary material is embodied in the tract in the ordinary hand of the commentary so as to be indistinguishable from the commentary at sight, and that the commentary itself occasionally points to the text by the use of such expressions as Cid ara tuc-somh (97), Cid ara n-ebairt (378, 484, 512, 385), intan roraidh (421), ata acht lem (2973), amal asbert i curp in libuir (173, 241) where corp in libuir always means the text of the book under comment.

Another but a rather uncertain criterion is this. A passage which does not occur by way of commentary on any previous quotation, but which is itself made the subject of commentary, is in a sense primary material, though not necessarily so old as the principal text on which the commentary is written.

The use of conagar is generally to introduce commentary even though the passage so introduced is itself subjected to comment. In a word, there is a primary commentary used to explain the original text, and a secondary commentary developing the content of the primary commentary (e.g., 1072 on 1068, 1637 on 1515). The etymological glosswork belongs to this last stage, and is incorporated without any regard to the context. The language even of the commentary is based on Old Irish usage. It explicitly recognises three genders in substantives and pronouns. In it airdíbdad (1264) means the silencing of the consonants, f and s. In later usage this term becomes airdibad, urdubad (uirdhiughadh, O'Molloy, Grammatica Latino-Hibernica 61), and denotes eclipsis, obnubilatio. The tract before us takes no account of eclipsis. At the time the tract was written the combinations mb, nd, had evidently not yet become assimilated (but cf. Nembroth, Nemruad). For, if such assimilations had taken place, an account would have been given of the phenomenon under such questions as What two consonants have the force of one consonant? (1375).

As regards ng initial, the evidence is not so clear. The nasal infection may have produced (ng+g) and not ng simply (255). On the other hand the combination is an Ogham letter (442)—but even vowels of diphthongs were pronounced separately (1430)—and is, considered along with the example, uingi (4926), curiously suggestive of: NT. N Latinum adiuncto Gamma Graeco significat semiunciam.Origg. xvi. 27, 4. The scheme of declension, also, distinguishes clearly between dative and accusative after prepositions (1651, 1770), a distinction not uniformly or often observed in Middle Irish, though a much later tract draws a distinction between acc. after a preposition importing motion, siubhal, and dat. after a preposition importing rest, comhnaidhe (Ériu, 8. 17, § 72, 73). This last, however, may be merely a grammatical recrudescence, or an imitation of Latin.

A few sporadic examples of Old Irish are here added:— 1. The article. nominative plural m. in muite 447, in taebomna tuissecha 918, in tri focail 2018, but ainm n., has art. nominative plural m. ind anmanda 4828. nominative plural f. inna iiii. aipgitri-sea 1132. For article developed from projected n., v. condelc, etargoire n-inchoisc 647, in incoisc 641.

2. NOUN Stems. A. o-stems: nominative plural n. araile crand 1149. B. io-stems: nominative singular n. a mberla sain 1044. dative singular oc nach ailiu 1044; accusative singular fria araill 3106, ar araill 5613; gan araill 3105. nominative singular &ersir; araill 3410; 'nas i n-aill 1272. quam i n-aill 4593, 4579 no da fhir-inaill 338. C. n-stem: gach reim n-olc 2177. 3. NUMERALS: teora, ceitheora 4708, 3747, cf. 872. 4. THE Verb: ailsius 5319, adrodamas 135; copula verb, arnid 693, nadat 4588.

As to the native elements, we are told that Cenn Faelad—in English Kinealy—wrote the Prologue (80). As this preface is not likely to have been omitted by the compilers of the extant tract, one concludes that this must be the actual introduction (1–62). This view is confirmed by the displacement in version ii. of the section (63–78) which is the work of a ommentator of Cenn Faelad; also by the particle tra in the first sentence quoted from Cenn Faelad, which follows the introduction in both versions.

There are four authors of the Auraicept proper, Cenn Faelad, Ferchertne, Amergen, and Fenius.

The excerpts from the Book of Cenn Faelad deal with: The origin of Gaelic (100). Divisions of the Latin alphabet (312), and of the Irish alphabet (392), Latin and Irish treatment of semivowels contrasted (445). Genders in Irish (520). Degrees of comparison in Latin, and qualitative and quantitative distinctions in Irish (639).

The excerpts from the Book of Ferchertne deal with: The seven elements of speech in Irish (739), and The formation and powers of Ogham letters (943).

There is a long excerpt from the Book of Amergen dealing with: the origin of Goedelg (1034). This passage is of earher date and language than the general run of the tract. In substance it is an alternative prologue.

The excerpts from the Book of Fenius (1102) deal with: The alphabets of Hebrew, Greek, and Latin (1129), hence probably the ascription to Fenius who was learned in those languages (160), and contemporary with the Exodus (1104).

Verse feet or syllabic content of Irish words (1213). Consonant changes (1264). The five kinds of Irish (1302). The twenty-five inflections (1515). What is alt? (1577). The end of the text of the Auraicept is noted (1636).

Besides those four ancient books cited, the Book of Cenn Faelad, the Book of Ferchertne (735), the Book of Amergen (1028), the Book of Fenius, Iair mac Nema, and Gaedel mac Ethiuir (1102), two others are mentioned, the Dúile Feda (5416), of which the Ogham tract is perhaps an expansion, and the Cin Ollaman (1204, 4385) possibly an early form of the tract on Metrics. The quotations from the first four books are set forth as usual in large hand; but possibly other passages from them are embodied in the commentary in the normal hand. For wherever a passage in the commentary is afterwards explained in detail with the usual artificial etymologies, this is an indication that the passage probably belonged originally to the ancient text.

While the ascription of the Book of Cenn Faelad is probably genuine, the same cannot be said of the Books of Ferchertne, Amergen, and Fenius. The quotations may be from writings approximately of the time of Cenn Faelad, but of unknown authorship. A commentator (1019–1027) takes the view that the work of these authors were successive steps leading up to the grand consummation, the Trefhocul. By the statement also of a commentator that what is first according to book order was invented last, to wit, the Book of Cenn Faelad (66) may be meant that this author co-ordinated all the ancient material, and presented it as it now stands. This view is upheld by another commentator who says that Ferchertne composed the Auraicept but Cenn Faelad rewrote it, or copied it, along with the greater part of Scripture (2638).

There seems no reason to question the ascription of the Book of Cenn Faelad to the author of that name. He is a well authenticated person. He died A.D. 679. His pedigree is found in the genealogy of the Cenél n-Eogain. His poems, dealing to a large extent with the wars of his kinsfolk, the Northern Uí Néill, are quoted largely in the annals. The curious tradition about his brain of forgetfulness (77) had no doubt a foundation in fact. Possibly he got a good education in youth, but developed a brain of forgetfulness by turning from learning to soldiering. He certainly fought in the battle of Moira A.D. 637, where he was wounded. Returning again to civil life and his early pursuits, poetry, words, and reading (78), he laid the foundation of that reputation which as Cenn Faelad, the Learned he still enjoys (O'Curry, Lectures). His period as an author therefore extends over the forty-two years between the battle of Moira and his death, and quotations from him must take rank among the oldest dated specimens of the language. But he refers to still older Irish writers, augdair na n-Gaideal (79), who wrote on the subject of Irish grammar, or of Irish origins. He may refer to such works as the Irish Chronicon Eusebii (Ériu 7, 62) which came down to A.D. 609, and of which the lost portion at the beginning may well have contained the story of Fenius. Writing in 603, S. Columbanus refers to antiqui philosophi Hiberniae as experts in chronography. Thus that earlier than the seventh century a state of learning existed which was held in esteem by the writers of that century is proved, though the direct products of that earlier learning are no longer extant. If we assume Cenn Faelad to be really the author, and therefore that the Auraicept was begun about the middle of the seventh century, how did it happen that while the other Western nations were sunk in ignorance, the Irish enjoyed the light of learning? Zimmer (Sitzungsberichte der Königl.-Preussischen Akademie der Wissenschaften, Dec. 1910, p. 1049) quoting the passage in Aur. 1859–1876 puts the question with great force: Das sind die Elemente der Kasus- und Numeruslehre,, wie man sie als Teil des über viele lahre sich erstreckenden Studiums der irischen fili (Grammatiker, Metriker, Antiquare und professionsmässiger Dichter) in den nationalen Schulen Irlands traktierte, als Klemens der Ire an der Hofschule Karls des Grossen jungen Franken das abc beibrachte, als Dicuil in St Denis, Dungal in Pavia, Sedulius in Lüttich und Metz, Moengal in St Gallen, Johannes Scottus an der Hofschule Karls des Kahlen tätig waren; durch diesen Unterricht ist Cormac mac Cuilennáin gegangen (gest. 908), der nebenbei ganz austándige Kenntnis im Latein, Griechisch, Hebráisch, Altnordisch, Angelsächsisch und Kymrisch besass.

The high tide of learning at a very early period in ancient Ireland was beyond a doubt caused by the influx of learned men from the Continent. In his researches Zimmer came upon this passage: Huni, qui ex nephario concubitu progeniti sunt, scilicet demonum, postquam praeheunte caterva viam invenerunt per Meotides paludes, invaserunt Cothos quos nimium terruerunt ex improviso monstro quod in illis erat. Et ab his depopulatio totius imperii exordium sumpsit, quae ab Unis et Guandelis, Gotis et Alanis peracta est, sub quorum vastatione omnes sapientes cismarini fugam ceperunt, et in transmarinis, videlicet in Hibernia, et quocunque se receperunt maximum profectum sapientiae incolis iliarum regionum adhibuerunt.

The first part of this statement relating to the Huns is taken from Jordanis, who wrote about A.D. 550, and fixes approximately the date of the depopulation of the empire and the rush of learned men into Ireland. We may assume that the migration had already continued for a time before this account was written. The intercourse between Ireland and the continent was certainly kept up.

Three centuries later we have this testimony respecting the Natio Scottorum quibus consuetudo peregrinandi jam paene in naturam conversa est. Quid Hiberniam memorem, contempto pelagi discrimine, paene totam cum grege philosophorum ad littora nostra migrantem!—(Sitzungsberichte der Königl.-Preussischen Akademie der Wissenschaften, 1910, p. 1080).

Zimmer with great learning, breadth of view, and mastery of detail builds upon these facts a history at once picturesque and surprising.

Stated briefly his hypotheses amount to this. The exodus from Gaul to Ireland (A.D. 419–507) was caused by the Homoousian persecution. Aquitania and the modern Baskish territory suffered like other parts, and Ireland was then the only haven of orthodoxy. Among the refugees from that region was the fatuus homunculus who was so called by his fellow-countryman the deacon Ennodius (A.D. 473–521) but who called himself Virgilius Maro, Grammaticus. He found an asylum with a native prince as was the fashion for learned men in those days, settled, and taught grammar, nay more, gained for himself fame, recognition, and a place among the native poets, being in fact none other than Ferchertne fili.

The Auraicept bears abundant evidence of the influence of two Latin authors, Isidore and Maro. The latter Zimmer laboured to identify with Ferchertne fili. It can be shown that the Auraicept lends no support to this proposed identification. This Ferchertne fili (<reg orig="CZ" resp="BF">ZCP</reg> 3, 13) is described in the tract as a contemporary of Conchobar mac Nessa (736), who, whatever reckoning be adopted, lived somewhere about the second century (cf. AU 484). According to this chronology, therefore, the identification of Ferchertne fili and Maro would place the latter at least a couple of centuries before his known floruit. Again the matter treated of by Ferchertne fili—the seven elements of speech in Irish, and the formation and powers of Ogham letters—does not correspond to anything in Maro's pages. If it be proved also that, while Isidore's influence is felt chiefly in the earlier part of the Auraicept, Maro's influence is confined entirely to the later, Zimmer's main contention that Maro was Ferchertne fili cannot succeed. Several centuries lay between the inception of the Auraicept and its close. Maro's tract had a profound influence on the Auraicept, but none on its early stages. According to internal evidence Cenn Faelad wrote the part ascribed to him about the middle or second half of the seventh century. That is the superior limit. The inferior limit lies in the eleventh century, or perhaps the tenth, and is determined generally by two facts—(i) that the Auraicept is found in two families of MSS., the variations in which postulate many generations of scribes, and (2) the immense development which has taken place in the tract itself as it has advanced from crude statements to a prosody which is exceedingly complicated and difficult. But the argument does not rest entirely on general considerations.

The second text (3382) quotes native grammarians by name, Ua Bruic, Ua Coindi, Ua Coirill and Ua Finn (3391). They are named by their surnames (each being the acknowledged head of his family), a usage that is not found earlier than the tenth century, one of the earliest instances being that of Ua Ruairc, AU 953. Ua Coirill mentioned above may have been the professor of law and history, who died AU1083. Hence the Auraicept was not completed before the middle of the tenth century, perhaps not till towards the end of the eleventh, when Maro's influence is still in evidence.

Let us now look at some excerpts from the works of the two Latin authors, Isidore and Maro.

I. Isidore of Seville, who died A.D. 636. His Etymologiae or Origines in twenty books contain a vast amount of information of such a sort that one finds it impossible to resist the conclusion that the compilers of the Auraicept had this document before them. At least that Cenn Faelad and Isidore drew matter from a common source is a certainty, for the facts (or alleged facts) and the phraseology are the same.

If we keep in mind that Isidore died the year before the battle of Moira, and that after that event Cenn Faelad began and pursued his studies with such success that he was popularly supposed to forget nothing (so one may interpret the words), and if we remember further that there was a constant coming and going of learned men, and a steady exchange of books between the continent and Ireland, there is no inherent improbability in the supposition that Cenn Faelad assimilated some of his material from the Origines published perhaps some twenty years before. True, the name of Isidore does not occur in the Auraicept, but no more does that of Eusebius from whom he probably made extracts, nor that of Luccreth Mocu Chiara (Älteste irische Dichtung, p. 51), from whose poem the passage about the seventy-two races (Aur. 215–227) was certainly taken.

There being no difficulty as to date or the omission of a name, full weight may be allowed to any other considerations tending to connect the two authors. The following quotations from many books of the Origines show how much the Auraicept was indebted to that source both in general structure and in detail.

Some references demonstrate that the Irish and Ireland were not unfamiliar to Isidore, at least as an author: Horrent et male tecti cum latratoribus linguis Scotti. Origg. xix. 23, 6. Scotia idem et Hibernia proximae Brittaniae insula, spatio terrarum angustior, sed situ fecundior. Haec ab Africo in Boream porrigitur. Cujus partes priores Hiberiam et Cantabricum Oceanum intendunt, unde et Hibernia dicta: Scotia autem, quod ab Scotorum gentibus colitur, appellata. Origg. xiv. 6, 6.

Time, place, person, and cause of writing (Aur. 63, 735, 1029), define the general plan and treatment of a subject, and are usually found in the íntroduction to any serious work in Irish.

Iam vero in elocutionibus illud uti oportebit, ut res, locus, tempus, persona audientis efflagitat. Origg. ii. 16, 1.

The cradle of letters was in Achaia, or by projection of d from art., Dacia, or by early French pronunciation, Asia.

Ubi fuit Athenae civitas. Origg. xiv. 4, 10. Apud Eotenam (uel Athena) civitatem. Aur. 214. Fuit autem Isis regina Aegyptiorum, Inachis regis filia, quae de Graecia veniens Aegyptios litteras docuit. Origg. viii. 11, 84.

These sentences show that, unless the Biblical Accad was introduced from some other source, Achaia (251) was probably the original reading; but the possibility that Achaia lay in Maeotidis Paludibus (<reg orig="CZ" resp="BF">ZCP</reg> 10 126) must not be overlooked.

Namque omnium ferocissumi ad hoc tempus Achaei atque Tauri sunt, quod, quantum conjicio, locorum egestate rapto vivere coacti. Glossae Juvenalis (Sall. Fragmenta).

Authority, written authority, ugdaracht (131), perhaps includes the following authors of whom, however, only two, Moses and Hieronymus (q.v.), are mentioned by name: Moyses, Dares Phrygius, Herodotus, Pherecydes. Vnde Sallustius ex historia, Livius, Eusebius et Hieronymus ex annalibus et historia constant. Origg. i. 42; 44, 4.

What are the names of the seventy-two races from which the many languages were learnt? (215, 263): Gentes autem a quibus divisa est terra, quindecim sunt de Japhet, triginta et una de Cham, viginti et septem de Sem, quae fiunt septuaginta tres, vel potius, ut ratio declarat, septuaginta duae; totidemque linguae, quae per terras esse coeperunt, quaeque crescendo provincias et insulas impleverunt. Origg. ix. 2, 2.

In definition a bias existed towards the heptad or the octave, Aur. 639, 739. De septem liberalibus disciplinis. Grammatica dialectica, etc., Aur. 51.—Origg. i. 2, 1.

Occasionally individual words are closely defined: Materia inde dicitur omne lignum quod ex ea aliquid efficiatur. Origg. xix. 19, 4. Fid, Aur. 943, cf. later the use of adbar.

The importance of Hebrew is insisted on: Illa lingua quae ante diluvium omnium una fuit, quae Hebraea nuncupatur. Origg. xii. 1, 2.

The Hebrew language was in the world first and it will remain after doomsday (190). Item quaeritur qua lingua in futurum homines loquantur. Origg. ix. 1, 13.

The following passage explains why Gaelic was deemed a worldly speech (46), not being one of the three sacred tongues in which was written the superscription on the cross (165). Linguarum diversitas exorta est in aedificatione turris post diluvium. Nam priusquam superbia turris illius in diversos signorum sonos humanam divideret societatem, una omnium nationum lingua fuit, quae Hebraea vocatur. Initio autem quot gentes, tot linguae fuerunt, deinde plures gentes. Tres sunt autem linguae sacrae: Hebraea, Graeca, Latina quae toto orbe maxime excellunt. His enim tribus linguis super crucem Domini a Pilato fuit causa eius scripta. Origg. ix. 1–3.

The early Irish rhythmical alliterative poetry, e.g.— arnin arding [d]éd, forsail for fot fedair, dinin disail for gair gabhaidh (1546), extending up to and running into the eighth century, might almost be defined by the words: Huic adhaeret rythmus, qui non est certo fine moderatus, sed tamen rationabiliter ordinatis pedibus currit; qui Latine nihil aliud quam numerus dicitur. Origg. i. 39, 3.

A verse of dithyramb or metrical rhythm is to be measured by a breath of the poet, five words to each breath (930). Periodos autem longior esse non debet quam ut uno spiritu proferatur. Origg. ii. 18, 2.

Grammatical questions as to gender and comparison of adjectives find a like expression in Latin and Gaelic: Neutrum dictum quia nec hoc nec illud, id est nec masculinum nec femininum. Origg. i. 7, 28; Aur. 614.

Octo autem modis conparatio analogiae colligitur: id est qualitate, conparatione, genere, numero, figura, casu, extremitatibus similium syllabarum, et similitudine temporum. Origg. i. 28, i; Aur. 639.

Non est maius nisi ad minus referatur. Sic et parvum opponitur magno ita ut ipsud parvum ad magnum, cui opponitur, sit parvum. Origg. ii. 31, 4, 5; Aur. 676.

Inde Ponticus sinus amplissimus a tergo Maeotidis paludibus; quod mare ex multitudine fluminum dulcius quam cetera. Sallust, quoted by Priscian, Macrobius, Servius, and Origg. xiii. 16, 4.

in dulci aqua xii. 6, 56;

sive salsae sint sive dulces. xiii. 14, 1; Aur. 730.

Artificial etymologies carry their influence into the Gaelic text; vir is derived from vīres, mulier from mollities, fēmina from fēmur: Vir nuncupatus, quia maior in eo vis est quam in feminis. Origg. xi. 2, 17; Aur. 605. Mulier vero a mollitie, tanquam mollier, detracta littera vel mutata, appellata est mulier. Origg. xi. 2, 18; cf. femina de flescda no maithchnechas, Aur. 610.

Femora dicta sunt, quod ea parte a femina sexus viri discrepet. Sunt autem ab inguinibus usque ad genua. Femina autem per derivationem femorum partes sunt, quibus in equitando tergis equorum adhaeremus. Origg. xi, 1, 106.

Femina vero a partibus femorum dicta ubi sexus species a viro distinguitur. Origg. xi. 2, 24; Aur. 608.

Consonants, semi-vowels, and mutes are treated similarly in the Gaelic and the Latin texts: Et vocatae consonantes quia per se non sonant sed iunctis vocalibus consonant. Haec in duabus partibus dividuntur: in semivocalibus et in mutis. Semivocales dictas eo, quod quiddam semis de vocalibus habeant. Mutae autem dictae quia nisi subiectis sibi vocalibus nequaquam erumpunt. Origg. i. 4, 3, 4; cf. Aur. 358 et seq.; 367 et seq.; 468 et seq. Vnde et legitimae nominantur illa ratione, scilicet vel quod ab E vocali incipiunt et in mutum sonum desinunt, ut sunt consonantes, vel quod a suo sono incipiunt et in vocalem E desinunt ut sunt mutae. Origg. i. 4, 10; Aur. 488.

The active and the passive of verbs: Etargaire persainni i ngnim (651); i cessadh (653). In persona verbi agentis et patientis significatio est. Origg. i. 9, 1.

The Origines contain well-known quotations (and the above may be of this sort): Litterae autem dictae quasi legiterae, quod iter legentibus praestent, vel quod in legendo iterentur. Origg. i. 3, 3; Aur. 360.

Some quotations are hard to find elsewhere: Nam unum semen numeri esse, non numerum. Origg. iii. 3, 1; Aur. 688.

It is not time that is divided but our actions (93).

Nam tempus per se non intellegitur, nisi per actus humanos. Origg. v. 31, 9.

These references I have not found.—Aur. 464, 517, 728.

The foregoing quotations are found in the portion of the Auraicept attributed to Cenn Faelad. They occur not only in commentary but often in the structure of the composition. Hence the conclusion that Cenn Faelad had before him the Origines or a document based thereon, and closely resembling it, is amply justified.

The use of the Origines is continued in the Gaelic text, after the portion attributed to Cenn Faelad ends. In the latter part of the book occur also some few suggestions of Ogham.

What is known as nihilus, Aur. 970, 8, is thus explained: V quoque littera proinde interdum nihil est, quia alicubi nec vocalis nec consonans est, ut QVIS. Vocalis enim non est quia I sequitur; consonans non est quia Q praecedit. Ideoque quando nec vocalis, nec consonans est, sine dubio nihil est. Origg. i. 4, 8.

A quotation common in the grammarians is: Nisi enim nomen scieris, cognitio rerum perit. Origg. i. 7, 1; Aur. 1099.

A quotation not seen by me elsewhere: Lapis autem dictus quod laedat pedem. Origg. xvi. 3, 1; Aur. 3396.

OccasionaIly the Latin helps to decide the reading of the Gaelic text: Incorporalia, quia carent corpus; unde nec videri nec tangi possunt, ut veritas, iusticia. Origg. i. 7, 4; cf. Aur. 3238.

Occasionally the Gaelic is a running commentary on the Latin: Perspicuae voces sunt quae longius protrahuntur ita ut omnem inpleant continuo locum, sicut clangor tubarum (stocaireacht no cornaireacht, Aur. 1477). Subtiles voces (cronan no certan bec, 1474) sunt, quibus non est spiritus, qualis est infantium vel mulierum vel aegrotantium, sicut in nervis (intan is cruit, 1484). Quae enim subtilissimae cordae sunt, subtiles ac tenues sonos emittunt (intan as bindi is tuiu &ersir; is isliu ata na a n-aill, 1484). Pingues sunt voces, quando spiritus multus simul egreditur, sicut virorum (mod .i. mo od .i. od ceol intan is mascul 1470). Acuta vox tenuis, alta, sicut in cordis videmus (traethait na ciulu isli na ciuil arda 1477). Dura vox est, quae violenter emittit sonos sicut tonitruum, sicut incudis sonos, quotiens in durum malleus percuititur ferum (intan is torand no is crand 1479, tourand no caint 4575).

Caeca vox est, quae, mox emissa fuerit, conticescit, atque suffocata nequaquam longius producitur, sicut est in fictilibus (tae a ed intan is fod 1479, fouta 4578). Origg. iii. 20, 10–13.

Occasionally the Latin determines the interpretation of the Gaelic, the latter being an almost literal translation of the former: Superflui sunt, quorum partes simul ductae plenitudinem excedunt, ut puta duodenarius. Habet enim partes quinque: duodecimam, quod est unum; sextam, quod duo; quartam, quod tria; tertiam, quod quattuor; dimidiam, quod sex. Vnum enim et duo, et tria, et quattuor, et sex simul ducta xvi faciunt et longe a duodenario excedunt. Perfectus numerus est, qui suis partibus adinpletur, ut senarius; habet enim tres partes, sextam, tertiam, dimidiam: sexta eius unum est, tertia duo, dimidia tres. Haec partes in summam ductae, id est unum et duo et tria simul eundem consummant perficiuntque senarium. Origg. iii. 5, 9—11; Aur. 1443—1453.

Occasionally the Gaelic gives merely the gist of the Latin: Primum enim diem a Sole appellaverunt, qui princeps est omnium siderum, sicut et idem dies caput est cunctorum dierum. Secundum a Luna, quae Soli et splendore et magnitudine proxima est, et ex eo mutuat lumen. Tertium ab stella Martis quae Vesper vocatur. Quartum ab stella Mercurii, quam quidam candidum circulum dicunt. Quintum ab stella Iovis, quam Phaethontem aiunt. Sextum a Veneris stella, quam Luciferum asserunt, quae inter omnia sidera plus lucis habet. Septimus ab stella Saturni, quae sexto caelo locata triginta annis fertur explere cursum suum. Origg. v. 30, 5–7; Aur. 3531–9.

Titles of chapters or sections in the Origines appear as names of Ogham: De homine xi. 1. ^… daenogam 5709. De avibus xii. 7. ^… enogam 5692. Oppida nobilia xv. 1, 6. ^… dinnogam 5687. De aedificiis sacris xv. 4. ^… ceallogam 5702. De navibus xix. 1,1. ^… ogam n-eathrach 6132. De instrumentis rusticis xx. 14. ^… ogam tírda 5724. De coloribus xix. 17 ^… dathogam 5697.

Bible names suffer change in passing into the Gaelic text through the Latin transliteration: Nebuchadnezzar, Nabuchodonosar, Origg. v, 39, 18; Nabgodon, Aur, 127. Nimrod, Nembroth, Origg. vii. 6, 22; Neamruad, Aur. 112. Noah, Noe, Origg. vii. 6, 15; Nóe, Aur. 107.

Secrecy—the avowed purpose of Ogham—is outlined in a simple code similar to that which finds expression in Aur. 6011. Caesar quoque Augustus ad filium, quoniam, inquit, innumerabilia accidunt assidue quae scribi alterutro oporteat et esse secreta, habeamus inter nos notas si vis tales ut, cum aliquid notis scribendum erit, pro unaquaque littera scribamus sequentem hoc modo pro a b, pro b c, et deinceps eadem ratione ceteras; pro s autem redeundum erit ad duplex a a. Quidam etiam versis verbis scribunt. Origg. i. 25, 2.

This reference I have not found: Aur. 3244–8, but cf. Maro 24, 10–24.

II. Virgilius Maro, Grammaticus

The editor, Huemer, in his Praefatio, p. xi., after giving a list of blunders common to all the MSS. of Maro, concludes: Atque archetypum illud litteris scotticis scriptum fuerit necesse est, cum a et u, c et t, r et s, s et f, p et f, saepe permutatae videntur.

The conclusion is irresistible. Whether the scribe was himself perpetrating these blunders, or, as his editor thinks, merely copying them from others, the sources of Maro, as we know him, are Irish.

Meyer, in two lists (Sitzungsberichte der Kgl. Preuss. Akad. d. Wiss. July, December 1912), gives, from Maro's tract, a selection of forty-two names, which he considers to be of Celtic origin. They are as follows: Aemerius p. (22). Andrianus (173). Arca rex (15). Asp-orius (5). Assianus (173). Bi-entius (137). Breg-andus (162). Don (15, 30). Fassica f. (123). Gabr-itius (126). Galb-arius (163). Galb-ungus (10, 122, 133). Gal-irius (146). Gall-ienus (129). Gelb-idius (36). Gerg-esus (15). Glengus (122, 133). Gurg-ilius (173). Iuu-anus (54). Lap-idus (19). Lassius (107). Lato-mius (123). Lugenicus (162). Mart-ulis (92). Mitterius (114). Ninus (119). Oss-ius (163). Perrichius (163). Plastus (151). Prass-ius (61). Regulus (?) (133). Rigas f. Rigadis (122). Rithea Nini regis uxor (119). Sagillius Germanus (17). Samm-inius, Virgil's uncle (28). Sarbon (122). Sarr-icius (123). Saur-inus (28). Sedulus (138, 139). Senenus (138). Sulpita (24). Ursinus (90).

Further examination may shorten the list without seriously disturbing the contention that if Maro had no connection with Ireland, his circle of Irish friends was unaccountably large.

Sua apte (116, 11; 81, 4) has been recognised as an Irish-Latin hybrid, su-apte, which later came into common use in Irish Latin.

There is a sprinkling of the loci communes of Latin Grammar, e.g.—

Maro denies that Latinitas is derived from Latinus, preferring latitudo, p. 5, 6: Aur. 355. litera ab ipsis etiam cerae caracteribus usque ad quassorum compossitionem hosce ordines directat, p. 7, 10; Aur. 1756. syllabae monades senas literas transcendere non debent ut scrobs, Maro, p. 11, 7; Aur. 1229. Grama est litteraturae peruidatio, quae quasi quaedam totius lectionis semitula est unde et a peritis litera interpretatur legitera quod est legendi itinerarium. Maro, p. 19, 11; Aur. 1758.

A certain resemblance is discernible between Maro, 24, 10–23, and Aur. 3244, and between Nec aperte masculinum nec absolute dicitur esse feminum. Maro, 31, 13; Aur. 614. verbum est omne quod lingua profertur et voce. Maro, 88, 6, and Aur. 1924.

The device scinderatio fonorum, Maro declares (p. 76, 7), was resorted to in order to sharpen the wits, to adorn expression, and: tertia (causa) ne mystica quaeque, et quae solis gnaris pandi debent, passim ab infimis ac stultis facile repperiantur.Maro, (p. 76, 7) The same reason, here called tertia, is alleged for the invention of Ogham:— Co mbeth in bescna-sa ic lucht in eolais fo leth, sech lucht na tirdachta &ersir; na buicnechta,Aur. 5472.

One device consisted in breaking up a sentence into groups of letters, e.g.— RRR. SS. PP. MM. NT. EE. OO. A.V.I., i.e., spes Romanorum perit. Maro, 77, 12; cf. Aur. 3501–3.

Also, words may be broken up into syllables, and these again may be strewn about in the jingle of a so-called sentence, e.g.— sicut Lucanus edidit; ge. ves. ro. trum. quando. tum. a. fec. om. ni. libet aeuo, [which is thus explained, quandolibet vestrum gero omni aeuo affectum.] Maro, 77, 6.

Or in single words, e.g. nodo for dono, nesi for sine, germen for regnum.—Maro 78, 28. This process appears in Irish as delidin sillabacda, Aur. 5312.

Amans may be transformed into manas (Maro, 79, 4), heri into hrei, is into si (78, 31); atat into tata (79, 10), a process which is called delitin litterda, metathesis of letters, Aur. 5308.

A meaningless syllable or disyllable may be introduced into a word, e.g. naviga-be-re for navigare, b-u-onum for bonum (Maro, 78, 17); forti-osi-ter for fortiter, compt-os-e for compte (Maro, p. 70, 6). A meaningless disyllable so introduced into Gaelic is called condall, Aur. 5317.

The unstressed syllable following an accented syllable is sometimes dropped, e.g. rogassem, rogasse for rogauissem, rogauisse; rogarunt, rogarit for rogauerunt, rogauerit (Maro, 78, 10). In Gaelic poetry this is called cotut, Aur. 5287.

Still more does the influence of Maro emerge in the Trefhocul.

The name Trefhocul bears a resemblance to the heading of the chapter De trimodo dicendi genere, Orig. ii. 17 which may have suggested it. Similarly the twelve items composing the Trefhocul might have been originally suggested by the duodecim latinitates of Maro, p. 88, 22, e.g. lumbrosa, hoc est perlonga, cum pro uno usitato totus uersus scribitur, with perlonga, cf. (can) rofota, Aur. 5060; and for the matter, cf. Aur. 5943 where each letter besides being written is spelt. sincolla, hoc est perbreuis, uersa uice cum totus uersus usitatus in uno continetur fono. With perbreuis, cf. (can) rogair, Aur. 5059, and for the matter, cf. Aur. 1326. belsauia, hoc est peruersa, cum casus nominum modusque uerborum transmutat. With peruersa, cf. (can) chlóen. —Aur. 5057, 5086. spela, hoc est humillima, quae semper res terrenas loquitur, with humillima, cf. a irisel, used of an appended syllable, a. — Aur. 5079, 5346. polema, hoc est superna quae de superioribus tractat. With superna, cf. a irard, an appended syllable, aib.—Aur. 5078, 5341. Assena, hoc est notaria, quae una tantum littera pro toto sono contenta est, cf. Q for ceirt, Aur. 5816, and R for Ruis 5820.

These coincidences are too numerous to be accidental. Omitting other lesser similarities, to lay stress on which might be regarded as fanciful, we come to the solid ground of quotation, (Hereon Zimmer, not having the whole tract before him, could find no footing.) metrofia, hoc est intellectualis, ut dictantabat, id est principium; sade, id est iustitia; gno utilitas; bora, hoc est fortitudo; ter hoc est dualitas coniugalis; rfoph, hoc est ueneratio; brops, hoc est pietas; rihph, hoc est hilaritas; gal, hoc est regnum; fkal, hoc est religio; clitps, hoc est nobilitas; mymos, dignitas; fann, hoc est recognitio; ulio, hoc est honor; gabpal, hoc est obsequium; blaqth, hoc est lux solis; merc hoc est pluuia; pal, dies et nox; gatrb, hoc est pax; biun, hoc est aqua et ignis; spax, longeuitas.

With the exception of y and z, which may have been added from another source, the explanation following hoc est, id est, is in each case identical with that given in Aur. 4211–4223.

Perhaps more important than all is the following: De h autem hoc dicendum est, quod semper inspirat, nunc ad fortitudinem, nunc ad motationem tantum. Nam cum semiuocalem praecesserit f, solum sonum pariter motabunt ut hfascon et faciunt f pro hf, si uero mutam c uel t uel p, suum sonum non amittit ut hcorda, htronus, hpalanx, Maro, p. 10, 9–14.

This passage throws light on Aur. 432, 1264–1279. Bogad there means aspiration (and apparently on finals) ut cloch, both. It has also another meaning, fortitudo, influenced by Ogham usage, where B + H = P, thus supplying the P which is non-existent in pure Gaelic. Semigud, again, means lenition and apparently on initials, but on this point the examples are inconclusive (cf. beith mo hsuidhe CZ 10. 266). Here the aspirating H precedes the consonant it aspirates, and thus Maro and the Auraicept are at one.

The warrantable conclusions to be drawn from the facts are few but very important. Bigerro sermone clefabo (Maro 8, 13) in the speech of Bigorre, which Zimmer presses to show that Maro was a native of that district, though in the tenth century he is called Tolosanus, proves merely that Maro was more or less conversant with Baskish. He was acquainted with viro athensi, a man belonging to the town of Ate south of Limoges (Maro 141, 28). He mentions a Sibylla Carginiensis, belonging to the town of Carca, in the Department of the Iberian Bastitani (p. 48, 25), and he knew a great number of Irishmen. Except perhaps in the passage last quoted in which h ad fortitudinem may be compared with b cum aspiratione pro p ponitur (432, 2879), no connection is traceable between him and Ferchertne fili, whose work belongs to a much earlier period than the Trefhocul. The influence of Maro's book on Irish grammar is confined to the Trefhocul, the last stage of the growth of prosody. The Auraicept proper, of which Ferchertne fili was one of the authors, or one under whose name ancient material was incorporated, shows no trace of Maro's influence.

Interesting questions arise in the text itself, some of which need only be mentioned, e.g.:—

The so-called mutes l, n, r, pronounced el, en, er in Latin, le, ne, in Gaelic (490, 511, 2981).

The frequent absence of aspiration, or aspiration by omission, of f and s, ni aimser fota 1576, ae aiges (408,9).

The confusion owing to the distance of the gloss from its text; e.g., 1515 is glossed at 1637, 1533–5 at 1675, 1577 at 1686, 1579 at 1687, 1591 at 1692, 1609–14 at 1695.

The tendency of words and phrases like alt co fesear (827, 1686), and fogni (1336, 1871) to become technical terms.

The French pronunciation of Latin, sirqundimus (4125), sircuim (4132), sircumplex (4784), siicuitas (2531), resulting in important changes in Gaelic, isinn Asia (2571) for isind Achaidh (251).

The rhymes, some apparently without sense (806), some without metre (1546), and some in metre but obscured by glosses (253, 4360, 5932).

The etymological reconstructions:— co-fid for cubaid (1512), ciallabair for ceileabar (1594), fegait for fichit (4735), for fégait, sedhait (4737), segait (4739), co, hógfégad for cóic (1637), ré huamma for réim (1638), so-fis for seis (479), ae gnithi for aicned (501), suad for uad (495), conod miait (508) for conid muiti (495), dorrae for trá (573), smitai, smit ai (= aue) for smita (4649).

The constant modernising of the text:— ceithri gne (872), ceitheora gnee (3747), moosom for moam (658), lugusom for lugam (659), cinntechsom, cinntichu son (1258) for cinntechem (4368).

Syllable, the ultimate element of everything in Gaelic except gender (1457). Number, case, person, degree, tense, mood, are indicated by syllables, whereas there is no distinction of gender indicated in spelling; and mod, tod, traeth, secundum, quosdam is aurlond (1496) or leading word that indicates gender.

The ascription of the same poem to Colum Cille (938), and to Cormac (1596, 3867, 5351).

The repetition of the same passages 1487, 1502; cf. 2616, 2622, shows that the present text is made up from at least two versions which sometimes contained the same material in different order. Hence no doubt comes the disjointed character of many passages.

The following terms, however, are of importance in order to understand the text:—

The word inrocomraircnigsiomairne gives the key to the plan of inflection called filltigthi, prepositional cases (1515). These eight syllables are held to form one word. According to our present grammatical methods the basis or unity is the compound word of five syllables comroircnigsemmar. It is preceded by a relative pronoun an- and by an enclitic or pre-verb -ro-, and it is followed by an emphasising pronominal suffix -ni. But the native Irish grammarians regarded all these syllables as parts of one word, and the scribes wrote the whole as one word. In their opinion proclitics were not separate words, but rather filltigthi, inflections, of the accented word. Accordingly, they wrote frissinfer as one word, an inflected form of fer, and gave it a distinct technical name. This also explains how is fer (1529) comes into the scheme. Is was an unstressed proclitic, and as such was treated as part of the word following. They did not observe that is fer, a thúarascbáil, had already been dealt with under the head of fer, a ainmniugud; nor did they recognise identity of case and inflection in the words which they wrote lafer, frissinfer; fofiur, iarfiur (1525).

Classification of prepositions, or any explanation of infixed pronouns (653) was thus rendered unnecessary.

Another flexion is réim, which later means oblique case (786). Of this flexion there are three kinds outward, inward, and both combined: outward ut est, fer. There is no flexion in the word as it stands in the nom., but there may be flexion in the context, e.g., in the accus., lasin (bf)fer; fir is an inward flexion of fer; and in fer is capable of both, e.g., dond fiur.

Taebreim prosta -i- fadéin (795) is the side-flexion, i.e., the external flexion of mé, tú, etc.

Tréfhocul rhymes with glé-accur (2179), and hence has ē and &fdot; ^{(f dot)}. It means three words (2018), and the knowledge of its secret, i.e., probably how it came to be so named, is very hard, considering that already thirty-six words have been found comprised under its species in Irish (2021). Tréfhocul came to mean a collection of precepts for the correction of incorrect versification. For each of the twelve technical faults (anocht), there were two correctives, each having its technical name, one belonging to the class called sciath, the other to that called gnúis. Thus the whole system of correct versification would have been comprised under a set of mnemonics, each mnemonic consisting of three heads, the name of the fault and the name of its two correctives—in short, it was a three-word scheme, and accurately named Tréfhocul. The original scheme of two correctives for each error is commended (2010–3), and still adhered to in rudrach (2047), and in uathad fri hilar (2057). But later refinements led to overlapping in the application of the correctives. Hence we find in the poem that a particular fault may be corrected or avoided by having recourse to more than one device of each class, sciath or gnúis.

We read that the 24 helps are increased to 47 (2126). The first list (2035–2071) totals 48, not 47; the second list (2083–2118) totals 51; the third list (2131–2176) totals 48. The discrepancies may be purely scribal, and due to a misreading of the Roman numerals, e.g., iii. read as iv., ii. as v., a constant source of error.

What is the difference between the two kinds of corrective? It will be found that all those called scéith, except lugugud, the addition of a diminutive suffix, are purely artificial distortions of the words; whereas those called gnúisi, except cenn&fdot;ochrus túis, airichill, dechned, and dichned, are in accordance with strict grammatical usage e.g., the use metri causa of so&fdot;er instead of fer (sóerugud); do&fdot;er instead of fer (dóerugud); the addition of two proclitic syllables (lorga fuach), or of one syllable (dialt n-etarléme); the use of issé, issi, issed (urlonn insce), where they might be omitted, e.g.— issí ind ala gnúis dég dil, urlonn insci ria hairim, (where issí completes the number of syllables required but might be dispensed with, if the number were complete without it); the use of singular for plural (óen), e.g.— creid uaim féin, is fíor mo rann, my quatrain, meaning (all the quatrains of) my poem; the use of plural for singular (lán), e.g.— meni fhuilet (2198) = meni fuil.

There must be some distinction of ideas in the two terms. The gnúisi are or were originally the natural devices, and the scéith the artificial devices for avoiding metrical faults, and perhaps the words were adopted on that principle, gnúis being the natural part of the man on the outlook to ward off an enemy, sciath the artificial implement for the same purpose.

A similar touch of imagination emerges in regarding head and heart as being supports of man, the male being (1808, 4994), and the further refinements of lánomna and their gene, mated pairs and their progeny. In the original notion doubtless the distinction was based on gender, but that fact was forgotten, and among the examples are lánomna deime (4999), mated couples (mas. and fem. in grammar), belonging to dem (a thing which is neuter in nature). This usage is even extended to quantity, which is still more remote from the original idea of gender.

The same tendency to personification appears in the suggested distinction among forcomét, frecomét, and degcomét (1818); forcomét, defensive armour, as kneecap on knee; frecomét, armament of offence, as knuckles; and degcomét, that which protects by supplying life and vigour.

Ogham.

Ogham alphabet was not of Irish origin (388, 2771).

Vielleicht, says Zimmer, schon dem 2. Jahrhundert n. Chr. ist der Verkehr des keltisch-römischen Westgalliens mit seiner alten Kolonie, dem keltischen Irland, die Einführung des Ogamalphabets in Irland und die Sitte, dem Dahingeschiedenen aufrechtstehende Steinpfeiler zu errichten, zuzuschreiben (SPA., 8th Dec. 1910, p. 1096).

According to MacNeill (p. 335) the origin of the Ogham alphabet must be placed after the Roman conquest of Gaul, because prior thereto the Western Celts of the continent used the Greek alphabet, and Ogham is based on the Latin alphabet.

In our knowledge of written Gaelic, Ogham inscription bounds the horizon, and the identity in value of the Ogham symbols with later MS. tradition is clear, with a few exceptions.

B Group.

Oghamists are agreed that F, the third letter of the group, must be read as V in inscriptions.

H Group.

In the Kilkenny Arch. Journal, July 1874, p. 231, Mr G. M. Atkinson suggested that this group is named after the first five Gaelic numerals, haon, do, tri, ceathar, cuig. This suggestion, without touching on the origin of H, is open to the objection that óen in O.I. is used only in composition with a substantive; but in the meantime it furnishes a useful mnemonic, and, as it stands, it indicates a possible connection between this group and numerosa, No. IV. of the duodecim latinitates of Maro, p. 89,9.

The difficulty is with regard to H, the first letter of the group. According to Maro H has two powers, ad motationem and ad fortitudinem, distinctions which correspond to the values in the text: (1) H non est litera sed nota aspirationis (767), and (2) B cum aspiratione pro p ponitur (433).

There is no demonstrated instance of H occurring in any of the Ogham inscriptions, and the sign may have originally been devised to represent a consonant value which became rare or obsolete before the time of the extant inscriptions; and the first value of H was attracted to, and became identified with, the symbol when the letter became familiar through Latin sources.

An endeavour is here made to establish the second or Ogham value of H from the following considerations.

A stop sibilant existed in Gaelic (cf. Pedersen, Gr. §51), corresponding to Gaulish Ð, which is sometimes written S, e.g., Lat. i-uuenc-us, Ir. ó-ac, Cym. ieu-anc; also without c, Ir. óa, Cym. ieu. The sibilant representing i appears also in Ir. as s-ó, s-óu, s-ó-om. That this sound is represented by Ogham H is rendered probable by the occurrence of the form ihuinnéis, Lat. juvenes, Ériu, viii. 5.

But this sibilant sound is also written d, e.g., Tadg = Tasg-os, and probably r, e.g., do-bidc = -dibirc (cf. Brér Garad for Brég Garad g .s. of Brí Garad). This value following B would give the Ogham B + H = P.

Again the three Ogham accents are represented in the text by the letters d, s, n, (4800). At lines 430, 2877, however, are found the three supplementa written h, s, n, except that at line 2878 for s = forsail is written the Latin sign of length (T has a sign that may be meant for s), and a particular sign is substituted for n. This leaves a probability that here H has the same value as D.

Teora fuillti ind Uraicepto (430, 2877) seem to be the three supplementa (cf. Origg. i. 3, 6), not of the Ogham but of the Auraicept, that is, they are additions made to the Ogham orthographic system by the grammarians of the MS. tradition. If this limitation be correct, examples of supplementa need hardly be looked for in the ancient Oghams. No opinion on this point is obtainable from modern Oghamists; for the word forbaid is hardly known, and Oghamists have hitherto ignored it. The word, however, occurs with definitions and examples in the Book of Ferchertne (810, 3633) one of the oldest parts of the text, and some of its provisions are exemplified, e.g., n (of cenn) is not doubled in Ogham (439) e.g. QENVENDANI, &Pgr;&egr;&ngr;&ngr;&ogr;-&ogr;&ugr;&igr;&ngr;&dgr;&ogr;&sgr; (Pedersen, Gr. §357). On the other hand a large number of inscriptions contain double letters. While some of these, dd and s, may perhaps be accents as indicated in the Auraicept, others like cc (1358, 1825) and ll (4788) obviously are not. Rhys Pedersen (Gr. § 4), and others incline to think them signs of lenition.

M Group.

The fact that the third symbol has the effect of two letters ng proves nothing as to that combination (4925).

In Ogham inscriptions the letters, if they belong to different syllables, are written separately, Studies in Irish Epigraphy i. 49.

The fourth symbol is said to represent sr or str, and the examples Stru 247, 2562, Streulae 5690, Strannan 5795, seem sufficient to establish that sound. The other examples point to a rare or obsolete sound like English z, e.g., stmólach 5695, sréghuindeacht 5801, súst 5727, srorca 5700.

No authenticated instance of this symbol has been found in inscriptions.

A Group.

The simple vowels have the same order and value as in Latin.

In epigraphy no distinction of long and short vowels has hitherto been observed.

Ea or Diphthong Group

The first and the last symbols ea and ae are interchanged.

The doubling of each letter in the explanatory script (1143) shows that the symbols stand for long vowels as well as diphthongs. Examples are given of ē and ō (2873), of ē and e (1285).

The symbol for ī (1369) is also used for p (Irish Epigraphy ii. 83; cf. MacNeill, p. 335, 6) and for medial y.

The symbol for ae (1365,70) is also used for x, which is regarded as a double c.

Prof. MacAlister (Irish Epigraphy ii. 144–8) has called attention to an excellent example—perhaps two—of Nathair im Ceann (5821). Owing to his axiom that the Oghams were not Cryptograms (Irish Epigraphy i. 66), he is unwilling to allow that the B and H groups were consciously interchanged (ii. 26, 140). But this interchange is contemplated (Aur. p. 306, 42), and since the study of the Oghams was elementary work prescribed for junior students, the wonder is perhaps that so many of the epigraphs are in regular Ogham.The remainder of the introduction on p. liv–lvi contains a concordance between the two families of texts and has been omitted. The reader is referred to the PDF version available ion the Internet.