so much favoured by the Romans, where, on its eastern slope, just below the summit, sheltered from the sea winds and open to the morning sun, they, during their tenure, set their station.

Beyond the ancient British name Caer Werid, the green city or camp, we have no record that we can call history of this earliest fortress, but on the verge of the hill, to the north-west of the castle and church, there is a long fragment of vallum and ditch that curves with the form of the hill, which tradition points out as the site of the ancient city. The curved lines and the exposed site of this entrenchment suggest adaptations of Roman work with these earlier lines, and doubtless they were so adapted by that nation to their own military use, as adjuncts and outworks to their own defences.

From the curved trenches of the north-west side a fosse and vallum are carried down to the river, defending the ford, and enclosing a suburb to the walled station; such annex being a common adjunct to Roman castra, for the accommodation of the civilian inhabitants and camp followers. Such an annex existed on the site of Foregate, Chester, external to the walls. At Lancaster this area bears the significant name of the Old Town; it is nearly open ground.

The site is now a grassy swelling hill, whence is a wide view of the North Country and the Lake mountains in front, and the wide bay of Morecambe with its sands to the left, whose shores and hills are likewise studded with early camps and mediæval towers and castles, the sites of which are visible from Lancaster.

Little, so far as I have been able to gather, has yet been done to trace either the British site or the Roman plan of Lancaster. It is essential that some such attempt should be made in dealing with Lancaster Castle, as the wildest myths as to its

character and origin are current, and are gravely published in local histories and guide books, and other works that should be of greater authority, as if they were recognised and accepted facts. Possibly the explanations I try to offer, based though they are on only a small collection of facts, may tend to set in the right direction the architectural history of the castle. Even if they should not suffice to do this, I venture to think that the entire architectural history of the building stands in need of revision, if, indeed, it should not be wholly re-written. Before treating of the castle it is necessary, for the proper understanding of its position, to endeavour to recover the hitherto untraced limits of the Roman station.

THE LINES OF THE ROMAN STATION.

It is claimed for Lancaster Castle that large portions of the existing buildings are Roman. Beyond the probability that the builders of the castle may have used up the materials of the Roman station, and partly occupied the Roman site, it is pretty certain that it does not contain in situ a single fragment of original Roman construction. Let us endeavour to find some trace of the actual Roman site, and then, before detailing the splendid mediæval features of the pile, try to show the origin of this legend. The only remnant of the Roman castrum is a small fragment of wall that stands on the summit of the steeply scarped declivity, overlooking the site of the old bridge, called Wery Wall.

66

Leland says of this wall, which he erroneously takes for that of the Priory (Itin. v, p. 85)— The olde waul of the circuite of the priory "commith almost to Lune Bridge. Sum have "thereby supposed that it was a peace of a waule "of the Toune, but indeade I espiyed in no place

H

"the Toune was ever waullid."

Leland's sup

position that this wall on the east was the priory wall is controverted by the fact that it is spoken of in the foundation charter of Lancaster Priory, where one boundary of its land is given as extending a veteri muro," proving it to be a long-existing wall before the foundation of the priory.

Mr. Roper states, in his history of the castle, that a portion of the Wery Wall was formerly visible at Vicarage Lane. Buck's view of Lancaster, dated 1725, appears to give some length of the eastern face, with a summer-house built on the angle of the still existing fragment. It is numbered and given in a list at the foot of the plate as the Wery Wall. It is shown as a buttressed wall, stretching as far as the castle. Roman walls were seldom if ever buttressed, and it is probable that this may have been built at a later date on the stillrecognisable base of the Roman wall. It is thus recorded by Stukely in 1720 :—

"I found a great piece of the wall at the north-east, in the garden of Clement Townshend, and so to Mr. Harrison's summer-house, which stands upon it. It is made of the white stone of the country, and still very thick, though the facing on both sides is peeled off for the sake of the squared stone, which they used in building. A year or two ago a great parcel of it was destroyed with much labour. This reached quite to the Bridge Lane, and hung over the street, at the head of the precipice, in a dreadful manner. From the summer-house it went round the edge of the close, north of the church, and took in the whole circuit of the hill. The ditch on the outside of it is now to be seen. I suppose that it originally enclosed the whole top of the hill where the church and the castle stand, which is steep on all sides, and half enclosed by the river Lune, so that it is an excellent guard to this part of the coast."

Buck's view gives also a line running from the existing fragment to the schoolhouse, now removed, which appears to coincide with the line of the north wall; but the uncertainty of definition in the engraving makes it doubtful whether this represents a hedge or the ruin of a wall. The curved entrench