the kind is known. There is a place over against the harbour called Penrallt Ceibwr. That is all I could learn.

The name Ty Llwyd is now shared by three houses close to each other, two of which are dilapidated. My theory is, that the old Ty Llwyd has disappeared, and that the present cottages have shared the name as a convenient arrangement, or as a compliment to an older and a famous Ty Llwyd."

Thus far Mr. Griffith.

Now, with regard to the well at Moylgrove, Fenton says (p. 537):-"the alum well at Treriffydd from which the late ingenious but eccentric physician, Dr. Owen, had in project a plan of extracting alum; but, if practicable, it was never carried into execution. Whether impregnated with that or not, I cannot say; but, as a chalybeate, after undergoing a most perfect analysis, it has been pronounced inferior to none, for that property, but the Tunbridge water. I never saw such an appearance of crocus, its inseparable characteristic. The spring is enclosed with stone and mortar, and, about 6 yards below, its stream is diffused into a more capacious basin excavated for the convenience of bathing in it." From this, one judges that the well has nothing to do with alum, save in the mind of "the ingenious but eccentric physician, Dr. Owen." Apparently, the original name of the fountain has, under the influence of the ingenious Dr. Owen, and under that of the rationalising tendency of the folk, wavered from Ffynnon Alwm to Ffynnon Halen. "John Pentrevor" finds it to-day a kind of midway between these, viz., Ffynnon Alem. It is important to know what the name was before the ingenious Dr. Owen came on the spot. Can anyone supply this information from an old document, or some very old resident in the place? There can be little doubt that Ffynnon Alem is a holy well, the word Alem being in all probability a form of a man's name. Lewis, in his Topographical Dictionary of Wales (London, 1833), says, under "Moylgrove":"Within the limits of this parish is a well, the water of which is considered efficacious in several diseases." Unfortunately, he does not indicate whether he is referring to Ffynnon Alem or not. But, considering that the well is enclosed with stone, like Ffynnon Gappan in Llanllawer churchyard, I think it very likely that Ffynnon Alem is that referred to. The folk etymology seems to point to the old name having been Alem, or Alan, or Alun.

The path and cave in the side of Pwll y wrach, mentioned by "Pentrevor," are explainable from Fenton, who says (p. 538):"On one side of which, almost in reach of the tide, an attempt

had been made to discover coal: and some smutty stuff very like that substance had been dug up, but the adventurers did not find encouragement enough to persevere."

Ty Rhôs, Fishguard.

The following is Professor Rhys' criticism, being a verbatim copy of his letter to the Editor of the Pembroke County Guardian:

"A happy New Year to the readers of the Guardian, and especially to Mr. Wade-Evans, who has been fixing the locality of Porth Kerdin, and Mr. John Griffith, who has been helping him. They seem to me to have succeeded, and their letters are very suggestive, and may lead possibly to more identifications. Of course such a name as Pwll y Crochan could not help very much, as it is so common; and the ambiguity of one like Trwyn Llwyd destroys its force by itself, as it may mean merely a grey headland, or a headland called after a person named Llwyd. The case is the same with Ty Llwyd, which might be either Grey House or Llwyd's House. The decisive point in Mr. Evans's argument is his being able to show 'Mesur y Pair’ as a natural feature which could be readily described in that way. Then the minor arguments come in with considerable force not far from Mesur y Pair' you have the small creek for landing, with Ty Llwyd close by, and you can hardly conceive the coincidence being a mere accident. What may be regarded as an accident is the disappearance of the name Porth Kerdin, which seems also to have been not an uncommon one. As to the little river Awen's name, that sounds to me very like the pronunciation of the Irish word for river, namely, amhain,' the equivalent of our afon.' Then Ceibwr is practically the same name which occurs in the Book of Llan Dav as Kibor, for the district around Cardiff; so it looks as if the names were Afon (Amhain) Cibwr, and Aber Cibwr, and that the Irish word amhain' (whence possibly the present Awen) had taken on itself the functions of the complete name : that is all.

"It is interesting to have Fenton's statement that Moylgrove had formerly 200 acres of wood and forest, covering ground which was in his time, as he says, totally denuded; but I hope he had reasons beyond those supplied by the names, Trewyddel and Moylgrove. On this point one would like to know what Dr. Henry Owen has to say. In the meantime I return to Trewyddel and Ty Llwyd, which I accept as meaning Llwyd's House, that is, the house of the Llwyd of the Mabinogi of Manawyddan, called in the Culhwch Llwydeu, which I am

inclined to regard as incorrectly spelled Llwyddeu in Lady Charlotte Guest's translation. The name is rather peculiar in its termination eu, but there was Arthur's son Llacheu, and the 'Englynion of the Graves' have a man called Tawlogeu (p. 34B). There are also feminines, such as Epilieu and Tegau Eurfron. There are probably more, but those are the only ones which occur to me now. I am not sure, after all, that the termination eu of Llwydeu may not be something quite different and quite non-Welsh. It reminds me of the Trallwng stone, with a Latin genitive Cunocenn-i by the side of a Goidelic genitive (in Ogam) Cunacenniv-i. The name appears later in Welsh as Concenn, Cincenn, Cyngen.

"Mr. Evans invites me to translate the Llwyd englyn in the Black Book, p. 35A: I only wish I could, but I do not expect to improve on Chancellor Silvan Evans's version. However, I suggest the following, with very great deference :

'Bet llvid lledneis, igkemeis tir.
kin boed hir tuw y eis.
dygirchei tarv trin ino treis.'

'Llwyd the Courtly's grave in Cemais land;
Though long the growth of his shaft,

Trouble was to assail there the bull of battle.'

On this guess I may remark that I cannot fix on the shade of meaning to give the adjective lledneis: perhaps it should be elegant, or comely, or courteous; but I hardly think modest, delicate, or dapper, would have suited one whom the poet calls immediately afterwards a bull of battle.' I have ventured to treat eis as meaning a shaft or spear-that is one kind of signification established for it by Silvan Evans in his Geiriadur. Further, I have supposed the sort of spear affected by Llwyd to have had a very long shaft. From the unexpected use here of the imperfect tense, I gather that the conflict in which Llwyd fell had been preparing for some time, perhaps at the hands of Pryderi, or else of Manawyddan, or some other leader of the Llyr family.

"Before losing sight of the Black Book englyn, allow me to call attention to the spelling Kemeis, which shows that the modern Welsh spelling should be Cemais or Cemmais, or else Cemes if you like, only not Cemmaes: there is no maes, field or plain, in it, as some people seem to imagine. The word is derived from cam, crooked, bent, and refers to the bend of a coast line or of a river, and challenges comparison, perhaps, with the cambus occurring here and there in Scotland. In the Pembrokeshire case the name refers, I fancy, to the angle made.

What sort of a How does it look

by the coast line with the mouth of the Teifi. headland is Pen Cemais? Is it high or low? from the sea? To leave this question of the name, I may point out that the Mabinogion give no indication that Cemais was not a part of Dyfed: indirectly I think that of Pwyll does the contrary. The poet Cynddelw-his name is to be pronounced as a dissyllable with the accent on the first syllable-refers (Myv., vol. i, p. 228) to Dyfed as 'bro seithbeu dyued' that is the country of Dyfed's seven pagi; for 'beu' is the mutated form of what we write 'pau'-it might be put into French as 'cher pays,'-in our patriotic songHen Wlad fy Nhadau,' sometimes Nadau in Morgannwg, as the Pentrefwr can testify. Now the 'seithbeu' or 'septem pagi' of Cynddelw can only have meant the' Seith Cantref Dyued' mentioned more than once in the Mabinogion of Pwyll and Manawyddan, the seven original cantrefs of Dyved, before Pwyll's son, Pryderi, had added to them. The lists of the cantrefs of Wales, together with the commots composing them, vary very much in different manuscripts, and it would be a very useful piece of work if they were to be carefully studied and explained by a trained historian. The seven here in question, as given from the Red Book of Hergest in the Oxford edition of the Bruts, are the following: Emlyn, Kemeis, Cantref Wartha (Upper Cantref), Deugledyf, Pennbrwe, Pebideawc, and Ros. One would like to know how Emlyn and the Upper Cantref, and perhaps more, were filched away from Dyfed; but I expect that is a question already dealt with in Owen's Pembrokeshire. I understand that portions of my native county of Cardigan have also been absorbed by Carmarthenshire. That octopus seems to have been equal to anything, except keeping hold of Gower.

The poet Ab Gwilym calls Dyfed generally Bro yr Hud or Gwlad yr Hud, that is to say, the realm of enchantment, glamour and illusion, the story of which, reaching the French romancers, became the theme known as the Enchantments of Britain.' But I think Mr. Evans is warranted in laying his finger on Cemais as the part to which the glamour adhered most thickly : that seems to follow from the stories reproduced in my Celtic Folklore, though I did not perceive it myself. In fact, I did not see the wood for the trees; but the wood is in this case of some importance, for it helps one to understand the story, if one may suppose the whole or most of Cemais to have once been more or less of a forest. Such a forest would serve also to shelter men who landed from the sea in such creeks as Aber Ceibwr. Possibly Llwyd may have been one of them; at any rate, I am inclined to identify him with Liath mac Celtchar of Cualu, the

most comely of the Fairy Chiefs of Erin. Now the country called Cualu, genitive Cualann, covered a tract of Ireland reaching from Dublin down as far as Wicklow. For, besides other proofs, Dublin is found called 'Ath Cliath Cualann,' that is, the Hurdle Ford of Cualu (Revue Celtique, vol. xv, p. 455), and in the other direction St. Patrick on his mission to Ireland is said, in the ancient manuscript called the Book of Armagh, to have put in at Inbher Dea in the territories of Cualu (in regiones Coolennorum), and the mouth of the Dea is known to be the Vartry river, which empties itself into the sea near the town of Wicklow. That coast must have been fairly convenient for communication with Dyfed. Cualu becomes in Welsh Cwl, for which there is some sort of Welsh tradition--I cannot find my references just now-that it was the part of Ireland from which Matholwch came, who occupies such a great place in the Mabinogi of Branwen. Somewhere, too, in the county of Wicklow, perhaps within the limits of Cualu, was Esgair Oerfel, whither Arthur went to attack Twrch Trwyth and his Boars. After some fighting, they are represented crossing the sea, and landing at Porth Clais at the mouth of the river Alun, below St. David's. Arthur, following them closely, seems to have landed in the same place, but is said to have spent the night at Mynyw, which I suppose means St. David's. This story, whatever else it may mean, seems based on some tradition as to invasions on the opposite coasts of Wales and Ireland in early times; but whether they are to be regarded chiefly as invasions of Ireland from here, or the reverse, does not seem very clear.

"The Twrch Trwyth and his Boars suggest to me the ruling family of a tribe whose totem was the wild boar, whose tribesmen were called boars, and whose ancestress was Banba, the lady of the boars, from Irish banbh, a young pig,' Welsh banw, 'a young boar.' Banba is used by Irish poets merely as one of the names of ancient Ireland; and the story of the advent of the Milesian Celts relates how they found Ireland possessed by three kings, whose wives were Eriu, Banba, and Fodla. The first of these goes with our Iwerddon,' and belongs more especially to Munster and the south-west, where the ancient Iverni lived; and possibly Fodla is to be identified with Ulster. The story of Twrch Trwyth seems to me now to settle the position of Banba in the east of Ireland, where we have found Cualu, that is to say, in Leinster. It looks as if the three queens' names reflect a triple division of Ireland in very early times, a sort of division, in fact, at which you would arrive by spreading Munster, Leinster, and Ulster out, so as to cover jointly the whole of the island. There is nothing to connect the