a day in digging on the spot suggested, but in vain. However, he did not give up the search, but resumed it another day, when his efforts were crowned with success; and he invited Mr. Breese of Port Madoc and the present writer to see them, which we did early in the month of October. We found that one of the stones reads: IOVENALI FILI ETERNI HIC IACIT. The letters are rather rude and debased capitals, and some of them are to some extent imperfect, especially the T of jacit. The second letter is slightly effaced on the right side; but that it is not a c is certain, for it is, among other things, smaller than the other c's, and than the other letters generally, which is frequently the case with o and 1, but not with c. So the inexplicable form Icvenali is disposed of in favour of Jovenali, which must be a form of the Roman name Juvenal, and stands here for Jovenalis. In O. Welsh it took the form Jouanaul, which occurs in the Liber Landavensis, pp. 166, 259. The form to be expected was Jouenaul, the a of the second syllable being due, perhaps, to the name being associated with Jouan, the O. Welsh form of Ieuan, John. As to Etern-i, the name is written Etterni in one of the Clydai inscriptions, and survives as Edern. Possibly the village of Edern, or rather Llanedern, in Lleyn, was called after the father of the Jovenal mentioned on the stone in question. It is also written Edeyrn I find; but that is a mistake, as it is not a compound of teyrn, which in all instances retains the accent; but the pronunciation Edéyrn is, I believe, unheard of. The inscription on the other stone consists of a single name, VENDESETLI, in taller and altogether much finer letters than those on the stone just described. A scrap of paper, purporting to be one of Mr. Jones Parry's former readings of this stone, handed me before either he or I had a chance of seeing the stone last year, gives the inscription as making VINDESETLI. How another reading got the preference in the Archæologia I do not quite know. But there can be no doubt that the first vowel is E; the second one, on the other hand, is rather faint, and from some points of view looks like an I, but we agreed that E is the correct reading. Then as to the name Vendesetli, it is the same as Vennisetli. The two together enable us to guess when nd was reduced to nn; and the e in the one answering to i in the other is also en règle, as showing that both represent an obscure vowel, the name being accented, in all probability, Vende-sétli. Compare also Cunatami and Cunotami, Senomagli and Senemagli, Trenayusu and Trenegussi. In manuscript Welsh the obscure vowel in question disappears altogether, so the steps are precisely what one would expect. Further, Irish enables one to see that vend- stands for an earlier vind-, which occurs only once in our early inscriptions, and that as vinn, namely on the Gwytherin Stone, in the name Vinnemagli, which is identical with the later Vendumagli of the Llanillteyrn Stone. This last proves that nd continued to be written for some time after the pronunciation had become nn, and disposes of the difficulty as to which of them may be claimed to be represented by the later form Gwenfael. Now Vende-setli means Vendesētli, whereof sētl- is the correct antecedent of our modern hoedl, life, lifetime; and it serves to give one an idea as to when the Welsh changed s into h, and e into oi, oe. In the Cambro-British Saints (pp. 267, 268) the name appears as Gvennoedyl, and in the Iolo MSS. (p. 141) it is Gwynhoedl. In the Myvyrian Archaiology (p. 741) it is Gwynoedl, which also occurs, p. 426, as Gwynodl, where o for oe marks the passing of the accent from the ultima to the penultimate. The church bearing the name of Llangwynoedl or Llangwnodl is in the neighbourhood, and for my part I have very little doubt that the monument in question was the tombstone of the saint after whom that church is called. The mention of him in the Myv. Archaiology, p. 426a of Gee's edition, is to the following effect: "Gwynodl or Gwynoedyl, son of King Seithennin, from Maes Gwyddno, whose land was inundated by the sea. Llangwnodl in Lleyn." The children of Seithennin, who were ten in number, and included among them Gwynhoedl, are said in the Iolo MSS., p. 141, to have become saints at Bangor, on the Dee. There is a similar reference to some of them also on page 105, where the name of the one here concerned is incorrectly given as Geneddyl. Then the passages in the Cambro-British Saints, of which the correct reading has been procured by Mr. Breese, explicitly connect the brothers Gwennoedyl and Tutclut with Lleyn. Lastly, it would be difficult perhaps to determine whether Tir Gwyn, as the field used (according to Mr. Jones Parry) to be called, meant the white or sacred land, or the land of Gwyn, that is of Gwynhoedl. As to the adjective gwyn, feminine gwen, meaning not only white, but also blessed and the like, I may add that, believing its original form vind to be a nasalised extension of the vid we have in the Latin video and its congeners, I should suggest that originally it did not mean a colour at all, but rather seen, beheld, spectatus, and that Vendesetl or Gwynhoedl means vir spectata vita, just as Hoedloyw, that is, I suppose, Hoedl-loyw must have meant him of the brilliant life': he was brother to Gwynhoedl. Lastly, as the O. Welsh habit was to write r for rh, there can be no doubt that Hiroidil on the Gwnnws stone stands for Hirhoidil, or, as it might now be written, Hirhoedl; and I am rather inclined to regard the early inscriptional forms Evolengi and Euolenggi as involving Celtic equivalents, now lost, of avum and longus. It is needless to add the other names into which bev, biu, now byw, 'life, lifetime' (as in yn dy fyw, 'in thy lifetime') are found to enter, such as Bodibeve, Conbevi, later Cynfyw, Biuhearn, &c., or to dwell on the fact that the single name Vendesetli gives us a whole chapter on the history of phonetic decay in Welsh. Would that a few more such were found! J. RHYS. 145 WELSH VERSIFICATION. THE following summary of the laws of Welsh versification is offered to the notice of those who may wish, when reading Welsh poetry, to understand the broad principles of its structure. More it does not pretend to be. At the same time the writer, while generalising as far as he saw practicable the confusing multiplicity of definitions given by almost all writers on the subject (all in Welsh, so far as he knows), has attempted to embrace all really salient points. The word verse in the following pages is used to designate one metrical line only. 1. The structure of verses is, in Welsh, founded on assonance (cynghanedd), rhyme (odli), and the number of syllables in each verse (cyhydedd). 2. Sometimes short pieces of poetry have also each verse beginning with the same word (cymmeriad geiriol), or with the same letter (cymmeriad llythyrenol). In the latter are included all successions of verses which begin with a vowel. This characteristic, though formerly much used, is now but rarely employed, and needs, therefore, no illustration. 3. Assonance (cynghanedd) consists in the recurrence, in one part of a verse, of one or more consonants (cynghanedd groes) or syllables (cynghanedd sain), which also occur in a preceding part of the same verse. Such recurring letters are here termed the assonants of those which precede, and to which they answer. 4. A consonantal assonance (cynghanedd groes) consists of one or more consonants in the latter part of a verse recurring in the same order as the same consonants in the first part, but affected by different vowels. No intruding consonant is allowed between any two asso nants. 4TH SER., VOL. VIII. 10 In the first verse of this couplet there are but two assonants, while in the second there are four. (a) It is not necessary that all the consonants in the first part should have assonants in the second. Even one will suffice, the other intervening consonants, if there be any, being simply passed over unnoticed, as in the first verse of the above couplet. (b) In this assonance, the first consonant in the verse, except n, which may or may not have its assonant, must have an assonant in the second part. (c) The most perfect form of the consonantal assonance is that in which the two parts of the verse can be interchanged without violating either sense or 5. A syllabic assonance (cynghanedd sain) consists in the rhyming of any syllable except the last in the second part, with a syllable in the first part of a verse. E. g., (a) “Wylo wrth rodio yr ydwyf." 1 2 (b) "Ni chewch ddyben o'ch penyd, In addition to the syllabic assonance there are also in the verse generally one or more consonantal assonances, which are not, however, subject to the same restrictions as a proper consonantal assonance. 6. (a) The letter h, when it stands alone, is sometimes regarded as a simple aspirate. It does not then necessarily interfere with or take part in any assonance. |