side by side with the historical Cymry, not left, as the Pelasgi were, in two or three obscure corners, but spread over a large part of the British Isles. But we are not left, in this case, to a doubtful passage of an ancient writer, to decide between the identity and the diversity of the two races. It is a plain fact of present experience, that the Gwyddyl, although a kindred people, are not the same people with the Cymry. Nor is the name “Gwyddyl” a mere nickname imposed by the latter tribe, and therefore perhaps applied by them in former ages to different races. For it is the name by which they call themselves, and it is therefore hardịy conceivable that they could have borrowed it from their neighbours. Moreover, the traditions of the Cymry speak of the Gwyddyl not merely as a race distinct from themselves, but as occupying those very parts of the British Isles which are occupied by them at the present day. The fact, that the Gwyddyl continue to exist in our immediate neighbourhood, is totally ignored by Dr. Owen Pughe.

If it be said that I rest too much upon traditional evidence; I answer, that I am ready to eliminate traditional evidence from both sides of the equation. There will remain, on my side, the evidence of local nomenclature plus intrinsic probability, and on the other side, nothing.

a

I must notice one or two minor points, confirmatory of the view advanced in my former paper. In p. 18 I cited from a document, the authority of which I will not undertake to discuss, the genealogy of the Gaelic chiefs of Gwynedd. Among them occur the names, “Cathal” and “Cathbalug,” which afford at least a colour of probability to the document. For the Irish word Cath, identical with the Welsh Cad, enters very extensively into the composition of Irish names, as Cad does into Welsh names. But the evidence in favour of the document, and of its genuine Gaelic origin, is considerably increased by a curious tradition which evidently grew out of a misinterpretation

of the name.

“ Cath,” I say, in Irish, is identical with the Welsh “ Cad;" but in Welsh it means neither more nor less than a “cat.” Accordingly, the chieftain Cathbalug, figures in Welsh mythology as a cat, the Palug Cat, as he is called, and is described as the progeny of the sow of Coll ab Collfrewi. This Cathbalug “ became ultimately one of the three oppressors of Mona that were reared in it.” Another member of the triad, Daronwy, was also one of the Gael of Mona, who after the final conquest of the island by Caswallawn Law Hîr, raised an insurrection, and recovered temporary possession of the country In

p. 56 I attempted to show that South Wales was settled, and in a certain sense civilized, earlier than North Wales; and that the latter country was regarded as a sort of fairy-land by the inhabitants of the former.

“ It is clear that, to the inhabitants of the south, Gwynedd was at this time an unknown land. Their imagination filled it with giants, fairies, monsters, and magicians. The inhabitants exercised strange arts: they had cauldrons of like virtue with that which renewed the youth of Æson: a red dragon and a white were buried as the palladium of their metropolis. Among their monarchs was a veritable cat, the offspring of a wandering sow. Their chief philosopher was of gigantic stature, and sat on a mountain-peak to watch the stars. Their wizard-monarch, Gwydion, had the power of effecting the strangest metamorphoses. The simple peasant, dwelling on the shore of Dyfed, beheld across the sea those shadowy mountain summits pierce the air, guardians as it seemed of some unearthly region. Thence came the mist and storm; thence flashed aloft the northern streamers; thence rose through the silent sky the starry path of Gwydion.”

Now these representations are entirely founded upon passages in the Mabinogion ; and I have since observed that Mr. Stephens (an admirable authority on the subject) is of opinion that all the Mabinogion in which those passages occur were written in Dyfed. Before I conclude this desultory and supplemental

2 Stephens' Literature of the Kymry, p. 424. ARCH. CAMB., NEW SERIES, VOL. V.

2 N

paper, I will venture to repeat a remark which fell from me last year at Brecon. The Gwyddyl are placed by tradition 'in various parts of Wales. The great body extended to Anglesey, Caernarvonshire, Merioneth, and Cardiganshire, with the adjoining parts of Denbighshire, Montgomeryshire, and Radnorshire. Another detachment occupied the western headland of Pembrokeshire, -another the district between the Towy and the Tawe, namely, Gower and the adjacent commots,-another the present county of Brecknock. What does this leave in the possession of the Cymry? The comparatively plain and fertile districts of Powys on the north-east, Gwent and Morganwg on the south-east, and Dyfed on the south-west. Here, then, we have two races, one in possession of the less accessible and less eligible region, the other occupying the rich plains and valleys ! Which are the aborigines, and which are the invaders ? Which are the assailants, and which are the defenders? I think there can be but one answer.

The Gael, then, upon my view, were assailed on three sides ; and the Cymry must have radiated subsequently from three points; namely, the north-east, the southeast, and the south-west. The conquerors of the southwestern region probably came by sea, and their name, the Demetæ, the people of Dyfed, seems to point to a connexion with the Dumnonii, the people of Dyfnaint, on the opposite shore of the Bristol Channel, as well as with an invading race who appear under a similar name in the annals of Ireland. And here we may have the true historical origin of the three tribes, distinguished by three distinct dialects, which occupy the Principality at present, the people of Gwynedd in the north, those of Gwent and Morganwg in the south-east, and those of Dyfed in the south-west.

W. Basil JONES.

a

ROMAN REMAINS IN WALES.

No. III.

Although after the interval of so many centuries it may seem futile to argue concerning antecedent probabilities and improbabilities connected with the Roman occupation of Britain,-and particularly of BRITANNIA SECUNDA, the province to which the researches of our Association are more immediately confined, -yet considerations of this kind can hardly be eliminated from the pure archæological analysis to which we wish to submit the subject we treat of. Where much uncertainty prevails, speculation and antecedent argument will always find a place.

Thus it would not be at all a loss of time if some antiquary would endeavour to discover what was probably the geographical condition of BRITANNIA SECUNDA. (1.) at the first ingress of the Romans into it; (2.) at the period of their finally quitting it. Some light has been thrown already on this part of the subject by the Rev. John Jones, of Llanllyfni, in the pages of our Journal, , when he treated of the Segontiaci;' and afterwards of the agricultural state of Britain, in what is commonly termed the Druidic period. If any one would see what a brilliant chapter of national history may be written upon such a portion of it, he has only to refer to the introductory part of Stuart's Caledonia Romana, where, what we could wish to see done for Wales, has been so for Scotland, with rare eloquence and acuteness. It is a task which must be attempted before the Cambria Romana can be called complete; but it is sufficient here to remind members of its importance, and to encourage them to turn their attention towards it. The use which we would make of it at this outset of our inquiry, refers to the probability of this or that direction of any given line of Roman road; the eligibility of this or that site of any supposed Roman station. 1 Vol. ii. p. 99.

2 Vol. i. New Series, Supplement, p. 89.

а.

For instance, it is impossible to discuss the probability of the fourth Base-line, which is one of our archæological desiderata,—that which we would wish to extend all along the western coast from SEGONTIUM to MENAPIA, without taking into account the traditional and historical changes which have taken place along that deeply indented line of shores. If it be asked why did the Romans, when they set out from SEGONTIUM (Caernarvon), plunge into the most difficult and dangerous passes of the country, going by the base of Snowdon, the real Mons Eryri (not to be confounded with the thence derived name of HERIRI Mons), instead of keeping outside those mountains, still to the south-east, and coming down upon what is now called Tremadoc? We must remember that in those days the Great Æstuary,—the Traeth Mawr,extended right up to the southern opening of the Pass of Pont Aberglaslyn; and being full of wide and highly dangerous sandbanks and quicksands,-as the Traeth Bach, a little further on, still is,- offered an insurmountable obstacle to all speedy passage.

Again, when the question is raised, how did the Romans cross into Mona?can the statement of Tacitus be true that they, the pedites, partly swam across the Menai Strait? Any person, looking merely at the actual condition of that strait, would pronounce the traject impossible without the aid of boats. But then we should on the other hand remember the shifting nature of the sandbanks at and below Moel y Don, the traditionary place of the passage for the Romans, (the certain one for the Anglo-Normans under Edward I.), which would render a swimming over by no means improbable. At that period too, if there be any truth in old British tradition, the wide æstuary between Beaumaris and Aber, now the Lafan sands, was dry and most probably cultivated ground. Bearing this circumstance in mind we are able immediately to understand the nature of the encampment of Bryn Briton, at the south-west end of the town of Beaumaris, which bears all the appearance of a Roman fortification, and whence there is reason to believe that a