Page images
PDF
EPUB

paved road stretched across the island of Mona, straight for the Roman station, the remains of which are in the churchyard of Holyhead.

The exit of the roads from Deva (Chester) cannot be determined without adverting to the probable condition of the æstury of the Dee, and the great marsh of Saltney (Salt Island?) lying at its head, during part of the middle ages.

We shall have to revert to this particular locality by and by; we here adduce it as an instance of the necessity of taking natural and geographical circumstances, and even probabilities, into account in many epochs of our survey.

From what we know of the geographical configuration, the geological formation, and the natural features of Wales, we may conclude, without any great improbability, that when this part of the island was formed into the Roman province of BRITANNIA Secunda, the valleys and river courses, in the more level portions of the country, were almost impracticable marshes and forests. The vale of the Severn, for instance, was mostly blocked up with bogs and woods, though the existence of two large Roman camps or stations, at Caer-flos, near Montgomery, and Caersws, near Newton, shows that open cultivable land, probably rich meadows as they now are, extended in those spots along Sabrina's course. The Vale of Clwyd must have had a dense jungle running all down the middle, while the Morfa Rhuddlan, at its northern extremity, was, like the wild marsh beyond Marathon, in old Hellas, impassable to an invading army. A retrospective glance at the vales of the Conwy and the Dyfi (Llanrwst and Machynlleth), will aid greatly in sketching out the lines where researches for Roman roads may be most successfully carried on. The South Sea islander who, two thousand

years hence, is to come to Britain and hunt up the traces of long forgotten and only traditional railroads, will, if he is a good archæologist, go upon the dead-level principle. Whereas we, at the present day, who are thus trying to discover the vestiges of our Roman conquerors, should certainly adopt the principle of open and dry land, rather than that of the mere " linea recta tutissima."

So again with Roman stations and forts. Many an antiquary of former days has lost much time in trying to assign a Roman origin for what was in reality a British hill-fort; and even at the present moment too much uncertainty prevails as to the probable characteristics of Roman sites, among no small number of otherwise intelligent and discriminating observers.

It is very allowable to speculate upon the causes which may

have induced the Romans to choose such and such sites for their roads and stations; though speculations must never be mistaken for facts, nor ought to be considered more than as aids to positive observation. There is no archæological absurdity in raising, for example, the questions whether at Deva and SEGONTIUM British towns and ports existed before the Romans came? and whether their previous occupation, as well as natural advantages, did not invite the settlement and determine the choice of the invaders. It is a fair question, though perhaps almost hopelessly obscure.

Whether Deva were or were not a British station before the ingress of the great conquerors, we may be sure of this fact, viz., that as there is little or no water at Chester, except what is supplied by the Dee, (the modern canals of course being out of the question), and as the tide makes the water brackish not far below the city, the Romans, when they fixed themselves at Deva, found some means of utilizing the fresh water of the Dee, and not improbably dammed up the stream where the mills now exist, and have existed beyond all record.

At SEGONTIUM, the Romans had the double advantage of a good port for vessels of small draught, and of a dashing mountain stream bringing an over abundant supply of that element which they knew so well how to appreciate, though modern civic corporations seem to omit it from the list of the indispensable necessaries of existence.

The natural advantages of each of these localities, the extreme points of our northern or third Base-line of the survey, could not be overlooked by any people, whether British or Roman.

Deva to SEGONTIUM. To commence then with the determination of the northern line of road in Britannia Secunda, that from Deva to SEGONTIVM: we observe at once upon the map that between ConoVIUM (Caerhun) and Deva the line of road—whereon the disputed station of VARIS or VARÆ may have been situated-must have passed first over the highlands of Denbighshire, and then over the Clwydian range, or round their northern end. We can perhaps at once eliminate this latter supposition from our inquiry by considering that the road from CoNovivm, if it went round the northern end of the range, where Prestatyn is situated, must have crossed what was then a dreary marsh below St. Asaph and Rhuddlan; and that its length, caused by such a circuit, will by no means tally with any of the distances marked in the Iter Antonini. It is by no means improbable that the Romans had good lines of passage over and among most of the Flintshire hills,—they were established on the opposite, or Cheshire, shore of the Dee, close to its very mouth ; and it is likely that they ferried themselves across, though the actual course of the river, now much altered in its channel, deprives us of the means of conjecturing the spot. The learned and accurate Pennant was misled into the error of considering a tower of the sixteenth century, just above Whitford, to have been a Roman Pharos; in the same way as other towers near Diganwy have been erroneously assigned to a similar origin. On the other hand it is stated that traces of the Romans have been found in the Flintshire lead mines ; and we see no improbability in assigning the Sarn Hwlcin or Sarn Wilkin (leading in a direction from the Dee to the Clwyd, not quite parallel to Offa's Dyke) to that people, merely because the name Sarn commonly refers to a road constructed with more than ordinary

care.

а.

[ocr errors]

Flint, from being a quadrangular town, built on a geometrical model, with a fosse and agger all round it,-and from Roman coins having been found within its area, -has been by some set down as a Roman station. It is however of mediæval,—that is to say of Edwardian origin,-and like all the new towns built by that great monarch, was traced upon a strictly geometrical plan. The presence of Roman coins is not conclusive, for

probably the Roman coinage circulated among the inhabitants of these islands long after the Roman power had been totally forgotten. Though therefore it is by no means improbable, -we would rather suppose it very much the contrary,--that the Romans had good communications along both banks of the æstuary of the Dee, we do not think that the line of the Iter of Antoninus is to be found so near the coast. On the contrary, we think that it may be fairly looked for as crossing by one or other of the Clwydian passes, and then ranging through the upper lands till it rounded or crossed the marsh land at the head of the æstuary, and finally entered the walls of DEVA.

Upon this subject we subjoin the conjectures and opinions of one of our members Mr. W. Wynne Ffoulkes, who we hope will make the verification of this portion of the line his own special study.

“The course of the Roman road from Chester to Bodfari has often occupied my thoughts, and, in the localities intervening, I have in vain sought for its trace; I say in vain, for I have found nothing positive among them. The first difficulty that presents itself is the loss of its egress from Chester. I can find nothing which gives any clue to this; and next, we are not yet quite sure as to the site of Varis; though, if we can verify what the late Mr. Aneurin Owen has advanced, viz., that its traces are to be found near Pontruffydd, that settles the question at once. A third difficulty exists in the absence of those local indicia which so frequently commemorate the line of a Roman road. I, there

3 We have observed something very like the traces of a Roman road crossing the Mold branch line near its point of divergence from the Chester and Holyhead Railway. Have any of our readers noticed this ?—EDD. ARCH. CAMB.

fore, am induced to think that, as far as our present data go, the course of the road in question must be left principally to conjecture; and, as we are now systematically studying the subject, even conjectures may be useful. I always looked for the Roman road between Bodfari Northop and Hawarden, until I found all those vestiges of Roman habitation on the top of Moel Fenlli (described in the Archeologia Cambrensis, vol. I. New Series). They suggested to my mind a route by Buckley and Mold to Moel Fenlli, by the present course of the road in that direction, assuming that the Saltney Marsh must have been crossed by a causeway and bridges, in the manner said to have been adopted by Severus. From Moel Fenlli I am unable to suggest the line taken to Bodfari; I see nothing to betoken it now. The distance would suit the Itinerary : Chester to Mold, eleven miles; Mold to Moel Fenlli, six; from thence to Bodfari might be accomplished in seven, I should think; making a total of twentythree. Horsley would correct it to twenty-two. On Buckley Mountain is a small hill, adjacent to the road, called the Knowl, which is a term not uncommonly given to a fortified post, like the Welsh Gop or Cop, and Mold looks very much like a town built on the model of a Roman station, with its castle mound at its head, the whole town lying on the slope of the hill, a position so much sought after by the Romans. But, further than these slight indicia, I see nothing in local names, or in the antiquities of the places themselves, to induce us to look for the Roman road there.

“On the other side of the country, between Bodfari and Northop, there are several remains which induce me to think that the road passed along that line. And I will begin first with a conjecture which, I think, is not unreasonable, considering the state of the country between Chester and Bodfari, and the object the Romans would evidently have in invading it. They did so, no doubt, for the purpose of subduing it (not merely for the sake of conquest) for the security of their conquests already made. This, I think, seems evident from Tacitus. Having conquered so far, they found it necessary to reduce the Ordovices and Silures also; and a terrible country they had to invade, the general features and character of which were mountains intersected by marshes. In Doomsday Book it is said that Rhos and Rhuvonioc were so marshy as to be wholly unfit for the plough. I, therefore, am inclined to think that, in their invasion, they would have followed a known route, rather than have made one of their own; and that, having done so, the route might eventually have become a road. Now I think there are traces of an ancient British route, by Hawarden, Moel y Gaer, and Moel Crio, leadARCH. CAMB., NEW SERIES, VOL. V.

20

« PreviousContinue »